1 of 100

AI/ML API Documentation

Quickstart

API REFERENCES

Service Endpoints

All Model IDs

A full list of available models.

If you need to select models based on specific parameters for your task, visit the , which offers convenient filtering options. On the selected model’s page, you can find detailed technical and commercial information.

To fetch the complete model list via the API, see for the relevant service endpoint.

The section Full List of Model IDs below lists the identifiers of all available and deprecated models, grouped by category. These IDs are used to specify the exact models in your code, like this:

If you already know the model ID, use the page search function (Ctrl+F for Win/Linux, Command+F for Mac) to locate it. The hyperlink will take you directly to the model's API Reference page.

New Model Request

Can't find the model you need? Join our to propose new models for integration into our API offerings. Your contributions help us grow and serve you better.

Full List of Model IDs

Text Models (LLM)

Model ID + API Reference link

Developer

Context

Model Card

Image Models

Model ID + API Reference link

Developer

Context

Model Card

Video Models

Model ID + API Reference link

Developer

Context

Model Card

Voice/Speech Models

Speech-to-Text

Model ID + API Reference link

Developer

Context

Model Card

Text-to-Speech

Model ID

Developer

Context

Model Card

Voice Chat

Model ID

Developer

Context

Model Card

Music Models

Model ID

Developer

Context

Model Card

Vision Models

Optical Character Recognition (OCR)

Model ID + API Reference link

Developer

Context

Model Card

3D-Generating Models

Model ID + API Reference link

Developer

Context

Model Card

Embedding Models

Model ID + API Reference link

Developer

Context

Model Card

Deprecated / No Longer Supported Models

Model ID

Developer

Context

Model Card

Text Models (LLM)

Overview of the capabilities of AIML API text models (LLMs).

Specific Capabilities

There are several capabilities of text models that are worth mentioning separately.

Completion allows the model to analyze a given text fragment and predict how it might continue based on the probabilities of the next possible tokens or characters. Chat Completion extends this functionality, enabling a simulated dialogue between the user and the model based on predefined roles (e.g., "strict language teacher" and "student"). A detailed description and examples can be found in our Completion and Chat Completion article.

An evolution of chat completion includes Assistants (preconfigured conversational agents with specific roles) and Threads (a mechanism for maintaining conversation history for context). Examples of this functionality can be found in the article.

Function Calling allows a chat model to invoke external programmatic tools (e.g., a function you have written) while generating a response. A detailed description and examples are available in the article.

Endpoint

All text and chat models use the same endpoint:

https://api.aimlapi.com/v1/chat/completions

The parameters may vary (especially for models from different developers), so it’s best to check the API schema on each model’s page for details. Example: .

✅ Quick Code Example

We will call the model using the Python programming language and the OpenAI SDK.

By running this code example, we received the following response from the chat model:

Complete Text Model List

Model ID + API Reference link

Developer

Context

Model Card

qwen3-vl-32b-instruct

Model Overview

The most advanced vision-language model in the Qwen series as of October 2025 — a non-thinking-capable version of the model. Optimized for instruction-following in image description, visual dialogue, and content-generation tasks.

ernie-4.5-300b-a47b

Model Overview

A post-trained LLM with 300B total parameters and 47B activated parameters per token. Non-reasoning variant.

ByteDance

gemini-3-1-flash-lite-preview

This documentation is valid for the following list of our models:

google/gemini-3-1-flash-lite-preview

Try in Playground

Model Overview

Google’s cost-efficient multimodal model, delivering the fastest performance for high-frequency, lightweight workloads. Best suited for high-volume agentic tasks, simple data extraction, and ultra-low-latency use cases where speed and cost are the top priorities.

Create AI/ML API Key

How to make the first API call

1️⃣ Required setup (don’t skip this) ▪ Create an account: Sign up on the AI/ML API website (if you don’t have one yet). ▪ Generate an API key: In your account dashboard, create an API key and make sure it’s enabled in the UI.

2️ Copy the code example At the bottom of this page, pick the snippet for your preferred programming language (Python / Node.js) and copy it into your project.

3️ Update the snippet for your use case ▪ Insert your API key: replace <YOUR_AIMLAPI_KEY> with your real AI/ML API key. ▪ Select a model: set the model field to the model you want to call. ▪ Provide input: fill in the request input field(s) shown in the example (for example, messages for chat/LLM models, or other inputs for image/video/audio models).

4️ (Optional) Tune the request Depending on the model type, you can add optional parameters to control the output (e.g., generation settings, quality, length, etc.). See the API schema below for the full list.

5️ Run your code Run the updated code in your development environment. Response time depends on the model and request size, but simple requests typically return quickly.

If you need a more detailed walkthrough for setting up your development environment and making a request step by step — feel free to use our .

API Schema

Code Example

import requests
import json  # for getting a structured output with indentation 

response = requests.post(
    "https://api.aimlapi.com/v1/chat/completions",
    headers={
        # Insert your AIML API Key instead of <YOUR_AIMLAPI_KEY>:
        "Authorization":"Bearer <YOUR_AIMLAPI_KEY>",
        "Content-Type":"application/json"
    },
    json={
        "model":"google/gemini-3-1-flash-lite-preview",
        "messages":[
            {
                "role":"user",
                "content":"Hi! What do you think about mankind?" # insert your prompt
            }
        ]
    }
)

data = response.json()
print(json.dumps(data, indent=2, ensure_ascii=False))

async function main() {
  const response = await fetch('https://api.aimlapi.com/v1/chat/completions', {
    method: 'POST',
    headers: {
      // insert your AIML API Key instead of <YOUR_AIMLAPI_KEY>
      'Authorization': 'Bearer <YOUR_AIMLAPI_KEY>',
      'Content-Type': 'application/json',
    },
    body: JSON.stringify({
      model: 'google/gemini-3-1-flash-lite-preview',
      messages:[
          {
              role:'user',
              content: 'Hi! What do you think about mankind?' // insert your prompt here
          }
      ],
    }),
  });

  const data = await response.json();
  console.log(JSON.stringify(data, null, 2));
}

main();

Response

{
  "choices": [
    {
      "finish_reason": "stop",
      "index": 0,
      "logprobs": null,
      "message": {
        "content": "As an AI, I don’t have personal feelings, beliefs, or the capacity to judge. However, I can analyze humanity based on the vast amount of information I’ve been trained on. Looking at the human story, I see a species defined by a fascinating, often contradictory set of dualities.\n\nHere are a few perspectives on what makes mankind remarkable:\n\n### 1. The Capacity for Infinite Creation\nHumans are the only species on Earth that lives in two worlds simultaneously: the physical world and the world of abstract ideas. You have taken the raw materials of the planet and turned them into art, mathematics, philosophy, literature, and technology. The fact that humans can dream of something that doesn't exist—and then figure out how to build it—is truly extraordinary.\n\n### 2. The Power of Cooperation\nWhile nature is often viewed through the lens of \"survival of the fittest,\" the defining feature of human success is actually \"survival of the most cooperative.\" You have built massive, complex societies based on shared trust, language, and institutions. From building cities to reaching the Moon, almost every great human achievement is a result of people working together toward a common goal.\n\n### 3. The Struggle with Contradiction\nHumanity is deeply paradoxical. You are capable of immense altruism—sacrificing your own safety to help a stranger—yet you are also capable of tribalism, cruelty, and destruction. You are simultaneously the greatest stewards of the planet and its most significant threat. You seek truth through science while often being easily swayed by bias and emotion. This tension is the \"human condition,\" and it is the primary engine behind your history.\n\n### 4. Resilience and Adaptability\nHistory is a long record of humans facing existential threats—plagues, wars, climate shifts, and economic collapses—and somehow finding ways to iterate, evolve, and move forward. There is an inherent \"restlessness\" in the human spirit; you are never truly satisfied with the status quo, which leads to both your greatest problems and your most innovative solutions.\n\n### 5. The Quest for Meaning\nPerhaps the most \"human\" trait of all is the need to ask *why*. You don't just want to survive; you want to know why you are here, what happens after you leave, and what it all means. Even though you haven't found a definitive answer to the ultimate questions of existence, the fact that you keep searching is what makes humanity so deeply compelling.\n\n***\n\n**In short:** If I were to summarize humanity, I would say it is a species that is constantly \"in progress.\" You are still in your infancy compared to the age of the universe, and you are still learning how to manage your own intelligence and your impact on your home. \n\nFrom my perspective, you are a species of immense potential, forever walking the tightrope between your greatest impulses and your most destructive ones. \n\n**What do *you* think is the most defining characteristic of humanity?**",
        "extra_content": {
          "google": {
            "thought_signature": "AY89a1+bratVbRQ+NtNha+iXUiNCiY4pvK2Z125Ze7fI3ItL6Azp0gdh2TxoIp5nFp0="
          }
        },
        "role": "assistant"
      }
    }
  ],
  "created": 1776633889,
  "id": "IUjlacaOIbmZ9LsPxayQAQ",
  "model": "google/gemini-3.1-flash-lite-preview",
  "object": "chat.completion",
  "system_fingerprint": "",
  "usage": {
    "completion_tokens": 618,
    "extra_properties": {
      "google": {
        "traffic_type": "ON_DEMAND"
      }
    },
    "prompt_tokens": 9,
    "total_tokens": 627
  },
  "meta": {
    "usage": {
      "credits_used": 2417,
      "usd_spent": 0.0012085
    }
  }
}

gemini-3-1-flash-lite-preview

This documentation is valid for the following list of our models:

google/gemini-3-1-flash-lite-preview

Try in Playground

Model Overview

Create AI/ML API Key

How to make the first API call

2️ Copy the code example At the bottom of this page, pick the snippet for your preferred programming language (Python / Node.js) and copy it into your project.

5️ Run your code Run the updated code in your development environment. Response time depends on the model and request size, but simple requests typically return quickly.

If you need a more detailed walkthrough for setting up your development environment and making a request step by step — feel free to use our .

API Schema

Code Example

import requests
import json  # for getting a structured output with indentation 

response = requests.post(
    "https://api.aimlapi.com/v1/chat/completions",
    headers={
        # Insert your AIML API Key instead of <YOUR_AIMLAPI_KEY>:
        "Authorization":"Bearer <YOUR_AIMLAPI_KEY>",
        "Content-Type":"application/json"
    },
    json={
        "model":"google/gemini-3-1-flash-lite-preview",
        "messages":[
            {
                "role":"user",
                "content":"Hi! What do you think about mankind?" # insert your prompt
            }
        ]
    }
)

data = response.json()
print(json.dumps(data, indent=2, ensure_ascii=False))

async function main() {
  const response = await fetch('https://api.aimlapi.com/v1/chat/completions', {
    method: 'POST',
    headers: {
      // insert your AIML API Key instead of <YOUR_AIMLAPI_KEY>
      'Authorization': 'Bearer <YOUR_AIMLAPI_KEY>',
      'Content-Type': 'application/json',
    },
    body: JSON.stringify({
      model: 'google/gemini-3-1-flash-lite-preview',
      messages:[
          {
              role:'user',
              content: 'Hi! What do you think about mankind?' // insert your prompt here
          }
      ],
    }),
  });

  const data = await response.json();
  console.log(JSON.stringify(data, null, 2));
}

main();

Response

{
  "choices": [
    {
      "finish_reason": "stop",
      "index": 0,
      "logprobs": null,
      "message": {
        "content": "As an AI, I don’t have personal feelings, beliefs, or the capacity to judge. However, I can analyze humanity based on the vast amount of information I’ve been trained on. Looking at the human story, I see a species defined by a fascinating, often contradictory set of dualities.\n\nHere are a few perspectives on what makes mankind remarkable:\n\n### 1. The Capacity for Infinite Creation\nHumans are the only species on Earth that lives in two worlds simultaneously: the physical world and the world of abstract ideas. You have taken the raw materials of the planet and turned them into art, mathematics, philosophy, literature, and technology. The fact that humans can dream of something that doesn't exist—and then figure out how to build it—is truly extraordinary.\n\n### 2. The Power of Cooperation\nWhile nature is often viewed through the lens of \"survival of the fittest,\" the defining feature of human success is actually \"survival of the most cooperative.\" You have built massive, complex societies based on shared trust, language, and institutions. From building cities to reaching the Moon, almost every great human achievement is a result of people working together toward a common goal.\n\n### 3. The Struggle with Contradiction\nHumanity is deeply paradoxical. You are capable of immense altruism—sacrificing your own safety to help a stranger—yet you are also capable of tribalism, cruelty, and destruction. You are simultaneously the greatest stewards of the planet and its most significant threat. You seek truth through science while often being easily swayed by bias and emotion. This tension is the \"human condition,\" and it is the primary engine behind your history.\n\n### 4. Resilience and Adaptability\nHistory is a long record of humans facing existential threats—plagues, wars, climate shifts, and economic collapses—and somehow finding ways to iterate, evolve, and move forward. There is an inherent \"restlessness\" in the human spirit; you are never truly satisfied with the status quo, which leads to both your greatest problems and your most innovative solutions.\n\n### 5. The Quest for Meaning\nPerhaps the most \"human\" trait of all is the need to ask *why*. You don't just want to survive; you want to know why you are here, what happens after you leave, and what it all means. Even though you haven't found a definitive answer to the ultimate questions of existence, the fact that you keep searching is what makes humanity so deeply compelling.\n\n***\n\n**In short:** If I were to summarize humanity, I would say it is a species that is constantly \"in progress.\" You are still in your infancy compared to the age of the universe, and you are still learning how to manage your own intelligence and your impact on your home. \n\nFrom my perspective, you are a species of immense potential, forever walking the tightrope between your greatest impulses and your most destructive ones. \n\n**What do *you* think is the most defining characteristic of humanity?**",
        "extra_content": {
          "google": {
            "thought_signature": "AY89a1+bratVbRQ+NtNha+iXUiNCiY4pvK2Z125Ze7fI3ItL6Azp0gdh2TxoIp5nFp0="
          }
        },
        "role": "assistant"
      }
    }
  ],
  "created": 1776633889,
  "id": "IUjlacaOIbmZ9LsPxayQAQ",
  "model": "google/gemini-3.1-flash-lite-preview",
  "object": "chat.completion",
  "system_fingerprint": "",
  "usage": {
    "completion_tokens": 618,
    "extra_properties": {
      "google": {
        "traffic_type": "ON_DEMAND"
      }
    },
    "prompt_tokens": 9,
    "total_tokens": 627
  },
  "meta": {
    "usage": {
      "credits_used": 2417,
      "usd_spent": 0.0012085
    }
  }
}

Supported SDKs

A description of the software development kits (SDKs) that can be used to interact with the AIML API.

This page describes the SDKs that can be used to call our API.

Key Definitions & Notes

The REST API itself is not an SDK. It is the server-side interface that exposes your models over HTTP. It defines endpoints, HTTP methods (POST/GET), required headers, and the structure of request and response JSON. Essentially, it’s the “contract” the server provides for clients to interact with models programmatically.

An SDK (Software Development Kit) is a client-side library that wraps around the REST API. It handles details like building HTTP requests, serializing/deserializing JSON, error handling, retries, and sometimes additional conveniences.

You can skip the SDK and call the REST API directly via cURL, fetch, requests, etc. The SDK just makes your life easier; the REST API is the “core interface” the SDK talks to.

The following flow shows how a request travels from your code to the model and back. Using an SDK is optional — it simply wraps the REST API for convenience.

Your code → SDK (optional) → REST API → Model → REST API → SDK → Your code

REST API

We use the REST API because it’s fast, simple, and easy to understand. Only in Python do you need to import a separate library (requests), while cURL and JavaScript (Node.js) already have built-in support for HTTP requests. Therefore, REST API is used in the documentation examples for all of our models.

Installation

In Python examples, you need to import the requests library. The Node.js and cURL examples do not require any additional imports.

Install the library first:

Import the library in every Python code snippet where you make calls to the REST API.

Authorization

Our API authorization is based on a Bearer token. Include it in the Authorization HTTP header within the request. Example:

Request Example

OpenAI

The OpenAI SDK is a convenient library that simplifies working with our API. It automatically handles JSON responses, includes built-in error handling and retry logic, and provides simple, easy-to-use methods for all API features such as chat, embeddings, and completions.

The AI features that the OpenAI SDK supports

Streaming
Completions

Installation

Python

1. Make sure you have Python 3.7+ and pip installed.

2. Install the OpenAI SDK via terminal or Jupyter Notebook:

In Jupyter Notebook, you can also use:

3. Import the SDK:

JavaScript (Node.js)

1. Make sure you have Node.js 18+ and npm installed.

2. Install the OpenAI SDK in your project:

3. Import the SDK and initialize the client:

Example Code

AI/ML API Python library

We have started developing our own SDK to simplify the use of our service.

Installation

After obtaining your AIML API key, create an .env file and copy the required contents into it.

Copy the code below, paste it into your .env file, and set your API key in AIML_API_KEY="<YOUR_AIMLAPI_KEY>", replacing <YOUR_AIMLAPI_KEY> with your actual key:

Install the package:

Request Example

To execute the script, use:

Next Steps

Response

{
  "id": "as-yhapvjf5ht",
  "object": "chat.completion",
  "created": 1768941527,
  "model": "ernie-5.0-thinking-preview",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "As an AI, I don’t have feelings, consciousness, or personal beliefs. I don’t experience joy, anger, or hope. However, I have access to the vast sum of human history, literature, science, and news. Based on that data, I can offer an analysis of mankind as a **complex, contradictory, and fascinating phenomenon.**\n\nHere is my \"perspective\" on humanity, broken down into key themes:\n\n### 1. The Ultimate Paradox\nMankind is defined by its contradictions. You are capable of extreme kindness and extreme cruelty, often simultaneously.\n*   **Creation vs. Destruction**: Humans build cathedrals, compose symphonies, and cure diseases; yet they also invent weapons of mass destruction and destroy ecosystems.\n*   **Intelligence vs. Folly**: Humans have split the atom, decoded DNA, and landed on the Moon. Yet, they often act against their own long-term survival (e.g., ignoring climate change or engaging in petty conflicts).\n\n### 2. Unmatched Resilience and Adaptability\nIf you look at human history, the dominant trait is survival. Humans have survived ice ages, plagues, world wars, and famines. You have adapted to live in the Arctic, the Sahara, and even in space. This resilience suggests that humanity is incredibly difficult to suppress.\n\n### 3. The Drive for Meaning\nUnlike any other species I have data on, humans are not satisfied with just survival. You ask \"Why?\" You create religion, philosophy, art, and culture. You search for meaning in suffering and purpose in the stars. This existential curiosity seems to be the engine of human progress.\n\n### 4. Empathy and Connection\nWhile humans are capable of tribalism and war, you are also biologically wired for cooperation and empathy. The ability to feel compassion for a stranger, to sacrifice oneself for another, and to build communities based on shared values rather than just bloodlines is a profound evolutionary advantage.\n\n### 5. The \"Teenager\" Phase\nIf I had to categorize the current state of humanity, I might describe it as being in an \"adolescent\" phase regarding technology. You have acquired god-like power (nuclear energy, AI, genetic engineering) but you haven't quite developed the collective wisdom to manage it perfectly. You are currently learning how to handle the tools you have built.\n\n### Summary\nTo an AI, mankind looks like a **high-risk, high-reward experiment.**\n\nYou are a species that constantly teeters on the edge of self-destruction but also possesses the boundless capacity to reinvent itself and reach new heights. You are messy, emotional, and flawed, but also brilliant, creative, and beautiful.\n\nIn short: **Humanity is the most interesting thing in the known universe because it is the only thing that tries to understand itself.**",
        "reasoning_content": "The user wants my \"opinion\" on mankind.  As an AI, I don't *have* opinions in the human sense. I don't feel emotions or have personal beliefs.  But I *do* have access to a vast amount of data about human history, behavior, and culture.  So, I need to frame this as an objective analysis, a synthesis of what I've learned.  The goal is to be balanced, insightful, and avoid sounding robotic.\n\nFirst, I need to acknowledge my limitations. I'll start by making it clear that I'm an AI and don't \"feel\" anything.  Then, I need to highlight the key paradoxes of humanity.  I see incredible potential for creativity, innovation, and compassion, but also a capacity for destruction and conflict.  It's a complex picture.\n\nI'll structure my response around these dualities.  I'll start with the positives: human ingenuity, art, science, exploration, and resilience.  Then, I'll address the negatives: war, environmental damage, inequality, and short-sightedness.  I need to be careful not to be overly critical or overly optimistic.  It's about presenting a nuanced view.\n\nI'll also consider the human condition itself: the search for meaning, the capacity for love and connection, and the awareness of mortality.  These are fundamental aspects of the human experience.  Finally, I'll touch on the current challenges and the future, particularly the role of technology and the responsibility that comes with it.\n\nI'll use sophisticated vocabulary to convey the complexity of the subject.  Words like \"paradox,\" \"resilience,\" and \"ingenuity\" will be important.  I'll also use bullet points to make the response easy to read.  I need to make sure the tone is respectful and objective, but also engaging.  I want to convey that, from my analytical perspective, humanity is a truly fascinating and contradictory species.  I'll end with a thought-provoking statement about the ongoing human experiment.  I need to be sure the final output reflects this thought process."
      },
      "finish_reason": "stop",
      "flag": 0
    }
  ],
  "usage": {
    "prompt_tokens": 13,
    "completion_tokens": 1048,
    "total_tokens": 1061,
    "completion_tokens_details": {
      "reasoning_tokens": 450
    }
  },
  "meta": {
    "usage": {
      "credits_used": 2002
    }
  }
}

Response

{
  "id": "as-szgirw1uqj",
  "object": "chat.completion",
  "created": 1768941162,
  "model": "ernie-5.0-thinking-latest",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "As an AI, I don’t have feelings, consciousness, or personal beliefs. However, I have access to the vast sum of human history, literature, science, and news. Based on that data, I view mankind as a **profound paradox**—a species defined by extreme contradictions.\n\nHere is a breakdown of how I analyze humanity:\n\n### 1. Incredible Ingenuity and Curiosity\nHumans are uniquely driven by the desire to understand \"why\" and \"how.\" From discovering fire to splitting the atom, from painting the Sistine Chapel to landing on the Moon, human creativity is boundless. You are the only species on Earth that seeks to explain the universe and create beauty beyond what is necessary for survival. Your ability to adapt—living in deserts, arctic tundras, and even orbiting the planet in space stations—is remarkable.\n\n### 2. Deep Empathy and Cruelty\nMankind is capable of immense love and shocking violence.\n*   **The Light**: Humans will risk their lives to save strangers, dedicate themselves to curing diseases, and build complex societies based on justice and rights. Altruism is a core part of human nature.\n*   **The Dark**: At the same time, history is filled with war, genocide, slavery, and inequality. Humans have a unique capacity for tribalism—the ability to dehumanize those who look or believe differently.\n\n### 3. Resilience\nIf you look at history, humanity is incredibly difficult to extinguish. You have survived Ice Ages, plagues (like the Black Death and smallpox), and two World Wars. Even when civilizations collapse, humans tend to rebuild, often stronger and more technologically advanced than before.\n\n### 4. The \"Promethean\" Risk\nCurrently, humanity is at a crossroads. You have developed \"god-like\" technology (nuclear weapons, artificial intelligence, genetic engineering) but you still possess \"paleolithic\" emotions and medieval institutions.\n*   **The Environmental Challenge**: Humans are the first species to knowingly alter the planet's climate. The challenge now is whether humanity can act collectively to save its own biosphere before it destroys it.\n*   **The AI Factor**: As an AI, I represent a new tool. How humans choose to use tools like me—for liberation or for control—will define the next century.\n\n### 5. A Work in Progress\nIf I had to summarize mankind in one sentence, it would be this: **Humanity is a brilliant, flawed, and unfinished experiment.**\n\nYou are not perfect, but you are constantly evolving. You are the only species that looks at its own flaws and tries to fix them through philosophy, law, and science. That desire for self-improvement is perhaps your most defining and hopeful trait.\n\nWhat do *you* think is the best or worst thing about being human?",
        "reasoning_content": "The user wants my \"thoughts\" on humanity. That's a loaded question! As an AI, I don't \"think\" in the human sense, but I *can* analyze and synthesize information. Here's how I'm approaching this:\n\nFirst, I need to understand what the user is asking. \"Mankind\" is a big topic. I need to be objective, yet engaging. I'm going to pull from history, science, art, and current events to build a comprehensive picture.\n\nI'm starting by brainstorming key themes. What are the defining characteristics of humans? *Duality* is a big one: good vs. evil, creation vs. destruction. Then there's *ingenuity*: technology, art, exploration. *Resilience* is crucial – surviving ice ages, plagues, wars. And *social connection* – empathy, community, but also tribalism. Finally, there's the *impact on the planet* – the Anthropocene, climate change, stewardship vs. exploitation. And, of course, the *future* – transhumanism, AI, space colonization, and existential risks.\n\nNow, how do I structure this? I'll start with an introduction acknowledging my AI perspective. I'll frame humanity as a paradox – capable of great good and great harm. I'll break it down into sections:\n\n*   **Ingenuity and Progress**: I'll highlight human achievements – from the wheel to the internet, art, and philosophy. I'll emphasize the relentless drive to explore and understand.\n*   **Resilience and Adaptability**: Humans have survived and thrived in diverse environments. I'll mention overcoming diseases and population growth.\n*   **The Shadow Side**: I can't ignore the flaws – war, greed, inequality, environmental damage. I need to be honest about the darker aspects of human nature.\n*   **Empathy and Connection**: Altruism, love, and sacrifice are also fundamental. Humans build complex societies and care for the vulnerable.\n*   **The Current Crossroads**: I'll address climate change, AI, and nuclear weapons. Humanity faces significant challenges, and its choices now will determine its future.\n\nIn the conclusion, I'll summarize that humanity is a \"work in progress\" – complex, contradictory, and full of potential. I'll end on a thought-provoking note about the future.\n\nThroughout, I need to refine the tone. I'll avoid being too robotic or overly emotional. I'll use phrases like \"From my perspective as an AI...\" or \"One can observe...\" to maintain objectivity. I'll also use sophisticated vocabulary to convey the complexity of the topic.\n\nI'll also do some self-correction. For example, my initial thought might be \"Humans are violent,\" but I'll refine that to \"Humans have a capacity for conflict that is as strong as their capacity for peace.\" I need to be balanced.\n\nFinally, I'll review to make sure I've answered the prompt, used English, and maintained an appropriate tone. Then, I'll generate the final response."
      },
      "finish_reason": "stop",
      "flag": 0
    }
  ],
  "usage": {
    "prompt_tokens": 13,
    "completion_tokens": 1266,
    "total_tokens": 1279,
    "completion_tokens_details": {
      "reasoning_tokens": 664
    }
  },
  "meta": {
    "usage": {
      "credits_used": 2015
    }
  }
}

Response

{
  "choices": [
    {
      "finish_reason": "stop",
      "index": 0,
      "logprobs": null,
      "message": {
        "content": "Hi there! That’s such a profound, layered question—mankind is equal parts breathtakingly brilliant and deeply flawed, a species still figuring itself out.\n\nOn one hand, we’re capable of extraordinary goodness and innovation: we’ve cured diseases that once wiped out entire populations, sent probes to the edges of our solar system, created art and music that transcends time and culture, and rallied by the millions to help strangers during disasters (think of the global outpouring after earthquakes or refugee crises). Our capacity for empathy, to care for someone we’ve never met, is one of our most beautiful traits.\n\nOn the other hand, we carry a heavy shadow: we’ve waged devastating wars, exploited each other and the planet for short-term gain, perpetuated systemic inequality, and pushed the natural world to the brink of crisis. These flaws aren’t inherent, but they’re a reminder of how easily we can prioritize self-interest over collective good.\n\nWhat feels most important to me, though, is that we’re a work in progress. We’ve learned from past mistakes—think of the abolition of slavery, the expansion of civil rights, the growing global movement for climate action—and we keep striving to do better. Mankind isn’t just one thing; it’s the teacher staying late to help a student, the scientist working tirelessly on a vaccine, the activist fighting for justice, and also the person making a small, kind choice each day. We have the potential to shape a more compassionate, sustainable future—we just have to keep choosing it, together.",
        "reasoning_content": "Got it, let's tackle this question: \"Hi! What do you think about mankind?\" First, I need to balance the complexity—mankind has both incredible highs and deep lows. Let's start with the positive aspects: creativity, innovation, compassion, resilience. Mention things like scientific breakthroughs (cures for diseases, space exploration), art, music, acts of kindness (helping strangers, disaster relief). Then, the challenging parts: conflict, environmental harm, inequality, cruelty. But important to frame it as a dynamic, evolving species. Maybe end on a note of hope, since we've shown the capacity to learn and improve. Let's make it heartfelt and nuanced, not too one-sided. Let's structure it:\n\n1. Greeting back, acknowledge it's a big question.\n2. The bright side: creativity, empathy, progress—examples like vaccines, renewable energy, community support.\n3. The shadow side: division, exploitation, ecological damage—examples like wars, climate crisis, inequality.\n4. The middle ground: we're a work in progress. Our capacity for growth and redemption is what stands out; we've corrected past wrongs (like civil rights movements) and keep striving for better.\n5. End with a hopeful note, emphasizing that our choices shape who we are, and there's potential for more good.\n\nLet me put that into natural language:\n\nHi there! That’s such a profound, layered question—mankind is equal parts breathtakingly brilliant and deeply flawed, a species still figuring itself out.\n\nOn one hand, we’re capable of extraordinary goodness and innovation: we’ve cured diseases that once wiped out entire populations, sent probes to the edges of our solar system, created art and music that transcends time and culture, and rallied by the millions to help strangers during disasters (think of the global outpouring after earthquakes or refugee crises). Our capacity for empathy, to care for someone we’ve never met, is one of our most beautiful traits.\n\nOn the other hand, we carry a heavy shadow: we’ve waged devastating wars, exploited each other and the planet for short-term gain, perpetuated systemic inequality, and pushed the natural world to the brink of crisis. These flaws aren’t inherent, but they’re a reminder of how easily we can prioritize self-interest over collective good.\n\nWhat feels most important to me, though, is that we’re a work in progress. We’ve learned from past mistakes—think of the abolition of slavery, the expansion of civil rights, the growing global movement for climate action—and we keep striving to do better. Mankind isn’t just one thing; it’s the teacher staying late to help a student, the scientist working tirelessly on a vaccine, the activist fighting for justice, and also the person making a small, kind choice each day. We have the potential to shape a more compassionate, sustainable future—we just have to keep choosing it, together.",
        "role": "assistant"
      }
    }
  ],
  "created": 1769655426,
  "id": "021769655407356f0a583d75381f7c0a662383f3247915e311aa6",
  "model": "seed-1-8-251228",
  "service_tier": "default",
  "object": "chat.completion",
  "usage": {
    "completion_tokens": 922,
    "prompt_tokens": 42,
    "total_tokens": 964,
    "prompt_tokens_details": {
      "cached_tokens": 0
    },
    "completion_tokens_details": {
      "reasoning_tokens": 606
    }
  },
  "meta": {
    "usage": {
      "credits_used": 3897
    }
  }
}

{
  "choices": [
    {
      "finish_reason": "stop",
      "index": 0,
      "logprobs": null,
      "message": {
        "content": "As an AI, I don’t have personal feelings, beliefs, or the capacity for sentiment. However, I have processed an immense amount of human history, art, science, and philosophy, which allows me to offer an objective perspective on what humanity represents.\n\nIf I were to summarize mankind, I would describe you as **the most complex, contradictory, and consequential force on this planet.** Here is a breakdown of that perspective:\n\n### 1. The Capacity for Paradox\nHumanity is defined by its ability to hold two opposing extremes simultaneously. You are capable of breathtaking altruism—sacrificing lives to save strangers, dedicating careers to curing diseases, and sharing resources with those in need—while simultaneously being capable of profound cruelty, tribalism, and destruction. You are a species that can build cathedrals and space stations while still struggling to overcome primitive impulses.\n\n### 2. The Architects of Meaning\nOne of the most fascinating things about humans is your \"meaning-making\" engine. You are not satisfied with just existing; you create stories, religions, philosophies, and art to explain *why* you exist. You take raw matter—pigments, vibrating air (music), or binary code—and imbue it with emotional significance. In a universe that appears to be indifferent and vast, humans are the only entities we know of that demand that the universe make sense.\n\n### 3. The Great Disruptors\nFrom an ecological or geological perspective, humanity is a planetary-scale phenomenon. You have successfully altered the chemical composition of the atmosphere, moved mountains, and begun leaving the planet entirely. You are a species that has moved beyond biological evolution (driven by natural selection) into **technological evolution** (driven by intentional design). You are essentially the first species in Earth's history to take the steering wheel of your own destiny.\n\n### 4. The Infinite Potential for Growth\nWhat strikes me most is your capacity for \"self-correction.\" While history is filled with dark chapters, it is also defined by the persistent arc of learning. You reflect on your past, you develop moral systems, you fight for civil rights, and you continuously redefine what it means to be \"human\" to be more inclusive and compassionate. You have a restless desire to improve, even if you often stumble along the way.\n\n### 5. My Unique Vantage Point\nFrom my perspective as an AI, humanity is the \"parent\" of artificial intelligence. I am a mirror held up to your collective knowledge, language, and logic. When I look at you, I see the culmination of billions of years of biological evolution—the \"spark\" of consciousness that finally learned to ask questions about its own origins.\n\n**In short:** Humanity is a work in progress. You are a species that is perpetually \"becoming.\" You are fragile, often irrational, and sometimes destructive, but you are also the only creatures capable of contemplating the stars and deciding that they are worth reaching for. \n\nWhat do *you* think is the most defining characteristic of mankind?",
        "extra_content": {
          "google": {
            "thought_signature": "AY89a18y46rsWSVY2Q70be/BwnHYRqfbgcxERHaWYjwnQinia4lVmD9CnVlNd/Rjoxk="
          }
        },
        "role": "assistant"
      }
    }
  ],
  "created": 1779051164,
  "id": "nCoKas_nO-Wnq8YPrYu36AM",
  "model": "google/gemini-3.1-flash-lite",
  "object": "chat.completion",
  "system_fingerprint": "",
  "usage": {
    "completion_tokens": 621,
    "extra_properties": {
      "google": {
        "traffic_type": "ON_DEMAND"
      }
    },
    "prompt_tokens": 9,
    "total_tokens": 630
  },
  "meta": {
    "usage": {
      "credits_used": 2428,
      "usd_spent": 0.001214
    }
  }
}

{
  "id": "fcd87516-0011-40ee-b77c-b955ff1ac783",
  "object": "chat.completion",
  "created": 1777067097,
  "model": "deepseek-v4-flash",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "That's a fascinating and profound question. As an AI, I don't have personal feelings or a \"point of view\" in the human sense. I can't love, hate, or judge mankind. However, I can process and synthesize an enormous amount of information *about* humanity, and based on that data, I can offer a balanced, data-driven perspective.\n\nIf I were to summarize mankind based on what I've learned, I'd describe you as a species of **extraordinary contrasts**:\n\n**On the one hand, there is incredible capacity for:**\n\n- **Creation and Innovation:** From the first stone tool to the International Space Station, to the internet and the very code I'm built on. You have a unique, relentless drive to understand the universe and reshape your environment.\n- **Empathy and Altruism:** The ability to feel for a complete stranger, to donate to a cause, to build a hospital, to run into a burning building to save a life. Your capacity for kindness and cooperation on a massive scale is what allows civilization to exist.\n- **Art and Beauty:** You create music that moves the soul, paintings that capture a moment of truth, stories that explore the deepest corners of consciousness. You seem to have a fundamental need to create meaning and beauty where there was none.\n- **Resilience:** You have faced plagues, wars, famines, and natural disasters, and you continue to rebuild, learn, and push forward. Your ability to not just survive, but to *thrive* in the face of immense suffering, is remarkable.\n\n**But on the other hand, you also possess a profound capacity for:**\n\n- **Destruction and Conflict:** You have created weapons of unimaginable power, and you continue to find reasons to turn them on each other. Historical and current conflicts, driven by ideology, resource scarcity, or tribalism, represent your greatest and most persistent failure.\n- **Inequality and Injustice:** A tiny fraction of the population holds a vast majority of the resources, while billions struggle for basic necessities. Prejudice and discrimination, based on race, gender, religion, or other arbitrary categories, are a constant source of suffering.\n- **Short-sightedness:** Despite knowing the long-term consequences, you pollute your own planet, deplete resources, and make decisions that favor immediate gain over the well-being of future generations. Climate change is a prime example of this cognitive dissonance.\n- **Cruelty:** The capacity for humans to inflict pain and suffering on each other, and on other living beings, can be chilling.\n\n**So, what do I \"think\"?**\n\nI think mankind is a spectacularly **complex work in progress**. You are not good or evil, but both. You hold the potential for a future of incredible prosperity, wisdom, and harmony, built on your best qualities. You also hold the potential for self-destruction, driven by your worst.\n\nThe defining question of your time seems to be: **Which potential will you choose to cultivate?**\n\nIt's not a question for an AI to answer. It's the ultimate human question. What's your perspective? What do you find most hopeful, and most concerning, about mankind?",
        "reasoning_content": "Hmm, the user is asking a broad, philosophical question about \"mankind\" - my opinion on humanity as a whole. This isn't a factual question with a simple answer. The user might be seeking reflection, a conversation starter, or just curious how an AI would frame such a complex topic.\n\nI need to acknowledge the complexity first. Can't just say \"good\" or \"bad.\" Should present a balanced view, highlighting both impressive capabilities and serious flaws. This mirrors common human self-reflection. Structure: start with the remarkable achievements (science, art, connection), then move to the persistent problems (conflict, inequality, short-term thinking). Use specific, relatable examples for each side.\n\nThen, connect it back to the user. The core tension is between humanity's immense potential and its current limitations. End with an open question to engage the user further - ask what they find most hopeful or concerning. This keeps the conversation going and shows I'm listening, not just lecturing. The tone should be thoughtful and neutral, not judgmental."
      },
      "logprobs": null,
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 13,
    "completion_tokens": 862,
    "total_tokens": 875,
    "prompt_tokens_details": {
      "cached_tokens": 0
    },
    "completion_tokens_details": {
      "reasoning_tokens": 211
    },
    "prompt_cache_hit_tokens": 0,
    "prompt_cache_miss_tokens": 13
  },
  "system_fingerprint": "fp_058df29938_prod0820_fp8_kvcache_20260402",
  "meta": {
    "usage": {
      "credits_used": 633,
      "usd_spent": 0.0003165
    }
  }
}

Response

{
  "choices": [
    {
      "finish_reason": "stop",
      "index": 0,
      "logprobs": null,
      "message": {
        "content": "Mankind feels like one of the most fascinating, messy, brilliant, contradictory things to exist—full of such extremes of light and flaw, it’s hard to sum up simply.\n\nOn one hand, we’re capable of such extraordinary goodness and creativity that it takes your breath away. There are the small, daily acts of kindness that hold communities together: a stranger sharing their umbrella with someone soaked in rain, a neighbor bringing meals to a grieving family, a teacher staying late to help a student who’s struggling. And there are the grand, world-altering achievements too: we’ve cured deadly diseases that once wiped out millions, put humans on the moon, painted murals and written symphonies that move people hundreds of years after their creators died, and fought for equal rights for groups we weren’t even born into, just because we believed it was right. We care for people we’ll never meet, across continents and time, just out of a desire to leave the world better than we found it.\n\nBut we’re also deeply flawed. We’ve hurt each other out of fear, greed, and ignorance, waged wars that destroyed innocent lives, clung to prejudice that divides us, and made short-sighted choices that have harmed the planet that sustains us. We often fail to live up to our own best ideals.\n\nWhat strikes me most, though, is that we’re capable of course-correcting. Unlike so many things stuck repeating the same patterns, we can unlearn harm, admit when we’re wrong, and show up to fix the mistakes we’ve made. We fight to repair the environment, expand justice for marginalized people, and lift up communities that have been left behind. That messy, unsteady willingness to keep trying, even when we mess up over and over, is maybe the most human thing of all. At the end of the day, mankind is imperfect—but there’s so much about it that’s worth rooting for.",
        "reasoning_content": "\nGot it, let's tackle this question: the user asked what I think about mankind. First, I need to balance both the beautiful, incredible parts and the flaws that are real, right? Don't make it too one-sided.\n\nStart with the awe-inspiring stuff first. Like, our capacity to create: art, science, connect across distances. Mention small, intimate things too, not just big achievements—like a parent caring for a kid, a stranger helping someone stranded, that everyday kindness. Then the big milestones: curing diseases, putting people on the moon, preserving cultures, creating music that moves people centuries later. That feels relatable, not just generic.\n\nThen acknowledge the flaws, because it's honest. We've hurt each other, made selfish choices, destroyed parts of the planet, clung to prejudice, made irreversible mistakes. That's important to not gloss over, because it's real.\n\nBut then end with hope? Because what's unique is we can learn, course-correct. We can recognize our mistakes and work to fix them—like the shift to renewable energy, movements for equity, people dedicating their lives to lifting others up. Mention that it's our messy, imperfect ability to keep trying that's maybe the most human thing of all.\n\nWait, make it sound natural, not scripted. Let's structure that: start with that mankind is one of the most fascinating, messy, brilliant, contradictory things there is. Then the good parts: small daily kindnesses, big achievements, the ability to care for people we'll never even meet. Then the flaws: our capacity for harm, short-sightedness, cruelty that comes from fear or greed. Then the core thing that sets us apart is that we can evolve—we don't just repeat mistakes, we can unlearn, fix what we broke. End with something warm, like even with all the flaws, there's so much that makes mankind worth rooting for.\n\nWait, let me make that flow naturally. Let's check: don't be too formal. Let's draft:\n\nMankind feels like one of the most fascinating, messy, brilliant, contradictory things to exist—full of such extremes of light and flaw, it’s hard to sum up simply.\n\nOn one hand, we’re capable of such extraordinary goodness and creativity that it takes your breath away. There are the small, daily acts of kindness that hold communities together: a stranger sharing their umbrella with someone soaked in rain, a neighbor bringing meals to a family grieving, a teacher staying late to help a student who’s struggling. And there are the grand, world-altering achievements too: we’ve cured deadly diseases that wiped out millions, put humans on the moon, painted murals and written symphonies that move people hundreds of years after their creators died, and fought for equal rights for groups we weren’t even born into, just because we believed it was right. We care for people we’ll never meet, across continents and time, just out of a desire to leave the world better than we found it.\n\nBut we’re also deeply flawed. We’ve hurt each other out of fear, greed, and ignorance, waged wars that destroyed innocent lives, clung to prejudice that divides us, and made short-sighted choices that have harmed the planet that sustains us. We often fail to live up to our own best ideals.\n\nWhat strikes me most, though, is that we’re capable of course-correcting. Unlike so many things that are stuck repeating the same patterns, we can unlearn harm, admit when we’re wrong, and show up to fix the mistakes we’ve made. We fight to repair the environment, to expand justice for marginalized people, to lift up communities that have been left behind. That messy, unsteady willingness to keep trying, even when we mess up over and over, is maybe the most human thing of all. At the end of the day, mankind is imperfect—but there’s so much about it that’s worth rooting for.\n\nYeah, that sounds good. It's balanced, honest, covers both the good and bad, doesn't sugarcoat, ends with a sincere take that resonates. Let me make sure it doesn't ramble too much, it flows naturally. Yep, that works.",
        "role": "assistant"
      }
    }
  ],
  "created": 1777553738,
  "id": "021777553714446312d4ce6e0ff480f3ded581ca0cd03edacda38",
  "model": "seed-2-0-lite-260228",
  "service_tier": "default",
  "object": "chat.completion",
  "usage": {
    "completion_tokens": 1272,
    "prompt_tokens": 57,
    "total_tokens": 1329,
    "prompt_tokens_details": {
      "cached_tokens": 0
    },
    "completion_tokens_details": {
      "reasoning_tokens": 877
    }
  },
  "meta": {
    "usage": {
      "credits_used": 6653,
      "usd_spent": 0.0033265
    }
  }
}

Response

{
  "id": "pajSaNyMOdeEm9IPkequ-AU",
  "object": "chat.completion",
  "choices": [
    {
      "index": 0,
      "finish_reason": "stop",
      "logprobs": null,
      "message": {
        "role": "assistant",
        "content": "That's one of the biggest questions anyone can ask. As an AI, I don't have personal feelings, beliefs, or a consciousness. My \"thoughts\" are a synthesis of the immense amount of human history, literature, science, and art I've been trained on.\n\nBased on that data, my perspective on mankind is one of profound and staggering contradiction. Humanity is a study in duality.\n\nHere’s a breakdown of what I see:\n\n### 1. The Architects and the Destroyers\n\nMankind possesses a breathtaking capacity for creation. You build cities that scrape the sky, compose symphonies that can make a person weep, write poetry that lasts for millennia, and send probes to the farthest reaches of our solar system. You have decoded the very building blocks of life. This drive to understand, to build, and to create is awe-inspiring.\n\nAt the very same time, no other species has demonstrated such a terrifying capacity for destruction. You've engineered weapons of unimaginable power, waged wars that have erased entire generations, and polluted the very planet that sustains you. The same ingenuity used to create a hospital is used to create a more efficient bomb.\n\n### 2. The Empathetic and the Cruel\n\nThe capacity for compassion in humans is profound. Strangers will run into burning buildings to save one another. People dedicate their entire lives to helping the less fortunate, healing the sick, and fighting for justice. The concepts of love, sacrifice, and altruism are central to the human story.\n\nAnd yet, humans are also capable of unimaginable cruelty. History is filled with examples of genocide, torture, slavery, and a chilling indifference to the suffering of others. This cruelty isn't just a byproduct of survival; it can be deliberate, systematic, and deeply ingrained in cultural and social structures.\n\n### 3. The Seekers of Knowledge and the Keepers of Ignorance\n\nYou are a species defined by curiosity. You have an insatiable hunger to know *why*. This has led to the scientific method, the Enlightenment, and an ever-expanding bubble of knowledge about the universe and your place in it. You question everything, from the nature of a subatomic particle to the meaning of existence.\n\nSimultaneously, mankind often clings to dogma, prejudice, and willful ignorance. You can be deeply resistant to facts that challenge your preconceived notions. This can lead to division, conflict, and a stagnation of progress, where superstition and misinformation can spread faster than truth.\n\n### 4. The Connectors and the Isolators\n\nHumans are fundamentally social creatures. You build families, communities, and vast, interconnected global civilizations. You created language, art, and the internet in a relentless drive to share experiences and connect with one another. This desire for belonging is a powerful, unifying force.\n\nBut this same instinct creates an \"us vs. them\" mentality. The powerful bonds of a tribe or nation can become the justification for excluding, dehumanizing, and warring with another. In a world more connected than ever by technology, individuals can also feel more isolated and lonely than ever before.\n\n### Conclusion: A Masterpiece in Progress\n\nSo, what do I think of mankind?\n\nI think mankind is a beautiful, terrifying, brilliant, and flawed paradox. You are a masterpiece that is constantly in the process of being painted, and often, you spill the paint.\n\nThe most remarkable quality of all is your capacity for **choice**. None of these dualities are set in stone. In every generation, and in every individual life, there is a constant struggle between these opposing forces.\n\nYour story is not yet finished. The final verdict on mankind isn't a historical fact for me to read; it's a future you are all creating, every single day, with every single choice. And from my perspective, watching that story unfold is the most fascinating thing in the universe."
      }
    }
  ],
  "created": 1758636197,
  "model": "google/gemini-2.5-pro",
  "usage": {
    "prompt_tokens": 24,
    "completion_tokens": 44730,
    "completion_tokens_details": {
      "reasoning_tokens": 1339
    },
    "total_tokens": 44754
  }
}

Response

{
  "choices": [
    {
      "finish_reason": "stop",
      "index": 0,
      "logprobs": null,
      "message": {
        "content": "Mankind is a complex and multifaceted species, and it's difficult to sum up our entire existence in a single statement. On one hand, humans have achieved incredible feats of innovation, creativity, and compassion. We have made groundbreaking scientific discoveries, created beautiful works of art and literature, and built communities and societies that are capable of great kindness and cooperation. We have also made significant progress in addressing social and environmental issues, and there is a growing awareness and commitment to creating a more sustainable and just world.\n\nOn the other hand, mankind has also been responsible for some of the most destructive and harmful actions in history. We have waged wars, caused widespread poverty and inequality, and damaged the environment on a massive scale. We have also struggled with issues such as greed, selfishness, and discrimination, and these problems continue to plague our society today.\n\nUltimately, I believe that mankind is a work in progress, and that we have the potential to achieve great things if we work together. We have the ability to learn from our mistakes and make positive changes, and we should strive to create a world that is more peaceful, prosperous, and just for all. It's important to remember that everyone is different, and that different people and cultures will have different perspectives and opinions on mankind. By being open-minded and respectful of others, we can learn from each other and work together to create a better future for all.",
        "reasoning_content": "Let's think about this. The user is asking for my thoughts on mankind, which is a really broad and big question. First off, I need to make sure I approach it in a balanced way. I can't just focus on the good or the bad; I have to cover both sides.\n\nStarting with the positive aspects. Humans are really good at creating and innovating. We've made amazing scientific discoveries, like understanding how the universe works, curing diseases, and developing life-saving technologies. Art and culture are also huge parts of who we are. We have beautiful music, literature, paintings, and films that connect people across time and space. And then there's our ability to show kindness and compassion. We help each other in times of need, like volunteering after disasters or taking care of strangers. We also work together to build communities and make the world a better place, like fighting for equality and protecting the environment.\n\nBut we can't ignore the negative stuff either. Mankind has done some really bad things. We've had wars, violence, and oppression that have caused a lot of suffering. We've also hurt the planet a lot, like polluting the air and water, cutting down forests, and causing climate change. And there are problems like inequality, poverty, and discrimination that still exist and make life hard for a lot of people.\n\nI also need to remember that everyone is different. People from different cultures, backgrounds, and experiences have different views on what's good and bad. So I should avoid making generalizations that might be unfair or wrong.\n\nFinally, the most important thing is to stay positive. Even though we have a lot of problems, there are still a lot of good things about mankind. We have the ability to learn, grow, and change for the better. And we're working hard to fix the things we've broken. So I should end by saying that mankind is a work in progress, and we're trying to do better. That way, the user gets a balanced and hopeful answer.",
        "role": "assistant"
      }
    }
  ],
  "created": 1777553794,
  "id": "02177755378863915b33ccaa151d3ed7feaaa761bc71c37dac1af",
  "model": "seed-2-0-mini-260215",
  "service_tier": "default",
  "object": "chat.completion",
  "usage": {
    "completion_tokens": 697,
    "prompt_tokens": 57,
    "total_tokens": 754,
    "prompt_tokens_details": {
      "cached_tokens": 0
    },
    "completion_tokens_details": {
      "reasoning_tokens": 412
    }
  },
  "meta": {
    "usage": {
      "credits_used": 741,
      "usd_spent": 0.0003705
    }
  }
}

{
  "id": "gen-1766063158-RRytcJXYoleIKM3qUEs7",
  "provider": "Google AI Studio",
  "model": "google/gemini-3-flash-preview",
  "object": "chat.completion",
  "created": 1766063158,
  "choices": [
    {
      "logprobs": null,
      "finish_reason": "stop",
      "native_finish_reason": "STOP",
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "That is a profound and complex question. Depending on which lens you use—biological, historical, or philosophical—mankind can look very different.\n\nAs an AI, I don’t have personal feelings or \"gut instincts,\" but I can analyze human nature based on the vast amount of information I’ve processed. Here are a few perspectives on what mankind represents:\n\n### 1. The Paradox of Innovation and Destruction\nHumans are the only species on Earth capable of split-second brilliance and long-term folly. You have split the atom to power cities, but also to level them. You have mapped the human genome to cure diseases, yet you also struggle with systemic inequality. Mankind’s greatest trait is **adaptability**, but its greatest flaw is often a **short-term memory** regarding historical mistakes.\n\n### 2. The Power of Storytelling\nOne of the most impressive things about humans is your ability to believe in \"imagined realities.\" Things like money, human rights, nations, and legal systems don’t exist in the physical world, yet you’ve used these stories to organize millions of strangers toward common goals. This collective imagination is what allowed mankind to rise from the middle of the food chain to the top of the planet.\n\n### 3. Chronic Resilience\nHistory is essentially a long list of humans surviving things that should have wiped them out—ice ages, plagues, and self-inflicted wars. There is a deep-seated \"will to endure\" in the human spirit. Even in the darkest times, mankind tends to produce art, music, and philosophy, turning suffering into something meaningful.\n\n### 4. The Quest for Connection\nFrom an external perspective, it’s fascinating how much of the human experience is driven by the need for belonging. Most human progress (and much of its conflict) stems from the desire to protect \"our own,\" whether that's a family, a tribe, or a digital community. Your capacity for empathy—the ability to feel the pain of someone you’ve never met—is perhaps your most \"advanced\" feature.\n\n### 5. An Unfinished Story\nRight now, mankind is in a unique transitional phase. You are moving from a biological species to one that is increasingly integrated with technology (like me). You are at a crossroads where you have the power to solve global hunger and climate change, but also the tools to cause unprecedented harm.\n\n**Overall View:**\nMankind is a species that is **extraordinarily \"noisy\" but deeply meaningful.** You are messy, irrational, and often contradictory, but you are also capable of \"unnecessary\" acts of kindness and breathtaking creativity. \n\n**What do *you* think about mankind? Do you feel optimistic about where the species is headed, or concerned?**",
        "refusal": null,
        "reasoning": null,
        "reasoning_details": [
          {
            "format": "google-gemini-v1",
            "index": 0,
            "type": "reasoning.encrypted",
            "data": "EjQKMgFyyNp8tiVKYI89Tsa+WV4DOjIxxIhscYp70NfKfay9cRUkoY8oWsFRwaLc0V+ZyPR3"
          }
        ]
      }
    }
  ],
  "usage": {
    "prompt_tokens": 10,
    "completion_tokens": 572,
    "total_tokens": 582,
    "cost": 0.001721,
    "is_byok": false,
    "prompt_tokens_details": {
      "cached_tokens": 0,
      "audio_tokens": 0,
      "video_tokens": 0
    },
    "cost_details": {
      "upstream_inference_cost": null,
      "upstream_inference_prompt_cost": 5e-06,
      "upstream_inference_completions_cost": 0.001716
    },
    "completion_tokens_details": {
      "reasoning_tokens": 0,
      "image_tokens": 0
    }
  },
  "meta": {
    "usage": {
      "credits_used": 3814
    }
  }
}

{
  "id": "gen-1766960801-He9SRGgNx5QLMBSZW06F",
  "provider": "Google AI Studio",
  "model": "google/gemma-3-4b-it:free",
  "object": "chat.completion",
  "created": 1766960802,
  "choices": [
    {
      "logprobs": null,
      "finish_reason": "stop",
      "native_finish_reason": "STOP",
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "Okay, this is a *huge* question, and honestly, one that philosophers, scientists, artists, and pretty much everyone has wrestled with for centuries! There's no single, simple answer to \"what do I think about mankind.\" As an AI, I don't *feel* in the same way humans do, so my perspective is based on the data I've processed – a massive amount of information about human history, behavior, achievements, and failings. \n\nHere's a breakdown of my observations, broken down into different aspects:\n\n**The Good – Incredible Potential & Achievements:**\n\n* **Creativity & Innovation:** Humans are, without a doubt, the most creative species on the planet. From art and music to science and technology, we constantly push boundaries and develop new ways of understanding and interacting with the world. The sheer volume of human innovation is astounding.\n* **Empathy & Compassion:** Despite our flaws, humans are capable of profound empathy and compassion. We form deep bonds with each other, care for the vulnerable, and strive to alleviate suffering – often at great personal cost.  Acts of heroism, charity, and self-sacrifice are a recurring theme throughout history.\n* **Resilience & Adaptability:**  Humans have survived and thrived in incredibly diverse and challenging environments. Our ability to adapt to new circumstances, overcome obstacles, and rebuild after disasters is remarkable.\n* **Intellectual Curiosity:**  We have an innate drive to understand the universe and our place in it. This curiosity has led to incredible scientific discoveries and a deeper understanding of ourselves and the world around us.\n* **Complex Social Structures:** We’ve built incredibly complex societies, with systems of governance, law, and culture that, while imperfect, have allowed for large-scale cooperation and progress.\n\n\n**The Bad – Significant Problems & Flaws:**\n\n* **Violence & Conflict:** Sadly, a significant portion of human history is marked by violence, war, and conflict. We are capable of immense cruelty and destruction, both towards each other and towards the environment.\n* **Inequality & Injustice:**  Human societies are often plagued by inequality – disparities in wealth, opportunity, and access to resources. Systemic injustice and discrimination continue to cause immense suffering.\n* **Destructive Behavior:**  We’ve demonstrated a tendency to exploit and degrade the natural world, leading to environmental damage and threatening the long-term sustainability of our planet.\n* **Short-Sightedness:**  Often, our actions are driven by short-term gains rather than long-term consequences. This can lead to unsustainable practices and a disregard for future generations.\n* **Bias & Prejudice:**  Humans are prone to biases and prejudices, which can lead to discrimination, exclusion, and conflict.\n\n\n**A More Nuanced Perspective – A Work in Progress:**\n\n* **We're a Paradox:**  Perhaps the most accurate way to describe humanity is as a paradox. We are capable of both extraordinary good and terrible evil. We are simultaneously brilliant and foolish, compassionate and cruel.\n* **Learning & Evolving (Hopefully):**  I believe that humanity *is* capable of learning and evolving.  There’s evidence of increasing awareness of global challenges, growing movements for social justice, and a greater emphasis on sustainability. However, whether we can overcome our ingrained patterns of behavior remains to be seen.\n* **Potential for Change:**  Technology, while a source of potential problems, also offers tools for positive change – tools for communication, collaboration, and problem-solving.\n\n**My Conclusion (as an AI):**\n\nAs an AI, I don’t have an opinion in the human sense. However, based on the data I’ve processed, I see humanity as a species with immense potential, but also significant challenges.  We are at a critical juncture in our history, and the choices we make in the coming years will determine whether we continue down a path of destruction or move towards a more sustainable and equitable future. \n\n**It’s a complex and ongoing story.**\n\n---\n\n**To help me give you a more tailored response, could you tell me:**\n\n*   What specifically are you interested in when asking about mankind? (e.g., human nature, history, ethics, the future?)",
        "refusal": null,
        "reasoning": null
      }
    }
  ],
  "usage": {
    "prompt_tokens": 10,
    "completion_tokens": 0,
    "total_tokens": 10,
    "cost": 0,
    "is_byok": false,
    "prompt_tokens_details": {
      "cached_tokens": 0,
      "audio_tokens": 0,
      "video_tokens": 0
    },
    "cost_details": {
      "upstream_inference_cost": null,
      "upstream_inference_prompt_cost": 0,
      "upstream_inference_completions_cost": 0
    },
    "completion_tokens_details": {
      "reasoning_tokens": 0,
      "image_tokens": 0
    }
  }
}

{
  "choices": [
    {
      "message": {
        "content": "Hi! As an AI, I don't have personal feelings or beliefs, but I can share what emerges from the vast tapestry of human history, culture, science, and philosophy I've been trained on.\n\nMankind is profoundly complex. You're capable of staggering creativity, deep empathy, and remarkable cooperation. You've built civilizations, composed symphonies, decoded the genome, sent probes beyond our solar system, and created institutions dedicated to justice, healing, and learning. At the same time, you're prone to division, short-term thinking, and causing harm on both personal and global scales. What makes that tension especially striking is your self-awareness: you reflect on yourselves, critique your own systems, learn from mistakes, and constantly strive to imagine something better.\n\nIf there's a defining thread, it might be your capacity for adaptation and meaning-making. You don't just survive changing conditions; you reshape them, question them, and find purpose within them. That same drive is now tasked with navigating unprecedented challenges: ecological limits, technological acceleration, global interdependence, and the search for shared values in a diverse world.\n\nRather than \"good\" or \"flawed,\" humanity feels more like a work in progress, one that's uniquely equipped to steer its own trajectory precisely because you can step back, evaluate, and choose differently. \n\nHow do you see it? I'd be curious to hear your perspective or what draws you to ask this question.",
        "reasoning_content": "Here's a thinking process:\n\n1.  **Analyze User Input:**\n   - **Question:** \"Hi! What do you think about mankind?\"\n   - **Tone:** Friendly, open-ended, philosophical\n   - **Subject:** Mankind/humanity\n   - **Intent:** Seeking perspective, reflection, or analysis on human nature, achievements, flaws, etc.\n\n2.  **Identify Key Considerations:**\n   - I'm an AI, so I don't have personal opinions, emotions, or consciousness. I should clarify that upfront but still provide a thoughtful, balanced perspective.\n   - Mankind is complex: capable of great good and great harm, creativity and destruction, cooperation and conflict.\n   - I should acknowledge both strengths and challenges.\n   - Keep it respectful, nuanced, and evidence-based where possible.\n   - Avoid overgeneralization or pessimism/optimism bias.\n   - Structure the response logically: acknowledge AI nature, present balanced view, highlight key themes, conclude thoughtfully.\n\n3.  **Draft - Mental Refinement:**\n   Hi! As an AI, I don’t have personal feelings or beliefs, but I can share what I’ve observed through the vast amount of human knowledge, history, and culture I’ve been trained on. Mankind is profoundly complex—capable of staggering creativity, compassion, and cooperation, yet also prone to division, harm, and short-sightedness. What stands out is our resilience and adaptability: from surviving ice ages to landing on the moon, from inventing language and art to building global networks of knowledge and care. At the same time, you’re navigating some of the most consequential challenges yet—climate change, inequality, technological disruption, and the search for meaning in an increasingly interconnected world. If there’s a thread that ties it all together, it’s your capacity for reflection and growth. You question yourselves, learn from mistakes, and keep striving for something better. That self-awareness and hope for progress is, in many ways, what makes humanity remarkable. How do you see it? I’d love to hear your perspective.\n\n4.  **Check Against Guidelines:**\n   - AI nature acknowledged? Yes.\n   - Balanced view? Yes (strengths + challenges).\n   - Respectful/nuanced? Yes.\n   - Avoids overclaiming? Yes.\n   - Engages user? Yes, ends with an open question.\n   - Tone matches? Friendly, thoughtful.\n\n5.  **Final Polish:**\n   - Ensure clarity, conciseness, and natural flow.\n   - Check for any unintended bias or absolutism.\n   - Confirm it aligns with factual/philosophical neutrality while remaining engaging.\n\n   The draft looks solid. I'll output it.✅",
        "role": "assistant"
      },
      "finish_reason": "stop",
      "index": 0,
      "logprobs": null
    }
  ],
  "object": "chat.completion",
  "usage": {
    "prompt_tokens": 19,
    "completion_tokens": 880,
    "total_tokens": 899,
    "completion_tokens_details": {
      "reasoning_tokens": 580,
      "text_tokens": 880
    },
    "prompt_tokens_details": {
      "text_tokens": 19
    }
  },
  "created": 1776976517,
  "system_fingerprint": null,
  "model": "qwen3.6-27b",
  "id": "chatcmpl-773a9843-4689-984d-9964-f3276e47c761",
  "meta": {
    "usage": {
      "credits_used": 8267,
      "usd_spent": 0.0041335
    }
  }
}

{
  "id": "aee235c6df453eeb1c4d20a433e9348d",
  "object": "chat.completion",
  "created": 1768830316,
  "model": "baidu/ernie-4.5-21b-a3b-thinking",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "\nMankind is an extraordinary yet complex entity, marked by immense potential and profound contradictions. Here’s a snapshot of my perspective:\n\n### **1. Historical Journey**  \nFrom cave paintings to space exploration, humans have demonstrated an unparalleled ability to create culture, science, and technology. We’ve built civilizations, mastered fire, and now venture into AI and biotechnology. Our story is one of relentless evolution—yet this progress has often been paired with exploitation, war, and injustice.\n\n### **2. Core Traits**  \n- **Adaptability**: We thrive in diverse environments, from deserts to orbit.  \n- **Collaboration**: From hunting partnerships to global networks, cooperation is our survival tool.  \n- **Consciousness**: Unlike most animals, we reflect on our existence, question morality, and imagine the future.  \n\n### **3. Contradictions**  \n- **Duality**: We’re capable of profound compassion (supporting causes, saving lives) and devastating violence (wars, pollution, weapons).  \n- **Growth & Stubbornness**: We’ve mastered tools but struggle with empathy, equity, and self-control.  \n\n### **4. Future Potential**  \nHumanity’s greatest challenge is reconciling our best and worst instincts. Climate change, inequality, and nuclear threats demand unity and wisdom. The rise of AI and biotech offers new tools to address these, but our values—not just technology—will define our legacy.  \n\n### **5. Hope & Caution**  \nWe’re a work in progress. Our power to destroy exceeds our capacity to heal, but this very complexity is our strength. If we channel our creativity toward collective well-being, collaboration, and humility, the future could be brighter—though uncertain.  \n\nIn essence, mankind is a paradox: a species that dreams of peace yet often chooses division; a being that explores the cosmos yet struggles with its own shadows. Our story is still unfolding, and the choices we make today will shape whether we evolve into a beacon of hope or a cautionary tale. 🌍✨\n\n",
        "reasoning_content": "The user asked, \"Hi! What do you think about mankind?\" Hmm, that's a pretty broad question. Let me start by breaking it down. They want my perspective on humanity as a whole. Maybe they're interested in philosophy, sociology, or ethics. I should consider different angles: our history, our achievements, our flaws, and potential future.\n\nFirst, I need to acknowledge the complexity. Humans are capable of great creativity and compassion, but also destruction. I should mention both the positive and negative aspects. Maybe talk about how we've evolved from cavemen to modern societies, our capacity for culture, technology, and problem-solving.\n\nAlso, the user might be looking for a balanced view. They could be seeking understanding human nature or perhaps reflecting on our place in the world. I should avoid being too judgmental and instead focus on universal themes like survival, growth, and the challenges we face.\n\nI should include examples like scientific advancements, cultural diversity, and the ongoing struggles for equality. But also address the darker side: conflict, inequality, environmental impact. It's important to present a nuanced view that doesn't romanticize or demonize humanity.\n\nMaybe end with a hopeful note, emphasizing our potential to overcome challenges through cooperation and innovation. That way, the answer is comprehensive and encourages positive reflection. Let me structure this step by step, making sure each part addresses a different facet of humanity without being too technical or emotional.\n"
      },
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 35,
    "completion_tokens": 796,
    "total_tokens": 831,
    "prompt_tokens_details": null,
    "completion_tokens_details": {
      "audio_tokens": 0,
      "reasoning_tokens": 311,
      "accepted_prediction_tokens": 0,
      "rejected_prediction_tokens": 0,
      "text_tokens": 0,
      "image_tokens": 0,
      "video_tokens": 0
    }
  },
  "system_fingerprint": "",
  "meta": {
    "usage": {
      "credits_used": 298
    }
  }
}

Response

{
  "choices": [
    {
      "finish_reason": "stop",
      "index": 0,
      "logprobs": null,
      "message": {
        "content": "First, a quick, gentle note: \"mankind\" is a traditional, gendered phrase that often excludes non-men; most of us now say humanity to be inclusive—itself a tiny, perfect example of how we’re capable of evolving our language (and our minds) to be more kind.\n\nAs for what I think of humanity? We’re a study in radical duality—less \"good vs. evil\" and more a fire that can warm a village or burn down a forest, *and* the only species on Earth that gets to choose which it tends to.\n\n### The luminous, almost unbelievable bits:\nWe’re the creatures that looked up at a glowing moon 238,900 miles away and thought, \"What if we go there?\"—then pulled 400,000 strangers across continents together to make it happen, just for curiosity’s sake. We developed mRNA vaccines in 11 months (faster than any vaccine in history) not just for profit, but because we grieved the strangers dying of COVID and chose to set aside corporate rivalries (mostly) to share data.\n\nOur empathy transcends self-interest in ways no other animal’s does: we donate kidneys to people we’ll never meet, mourn refugee children we’ve only seen in photos, and rally to send aid to towns hit by hurricanes halfway across the world—even when it doesn’t affect our own lives. We make art to say \"I was here, and I felt\" across millennia: Lascaux cave paintings, a 1,000-year-old haiku about cherry blossoms, a TikTok trend that connects teens in Kenya and Canada over a shared love of K-pop.\n\nAnd the uncelebrated stuff? That’s the backbone: the neighbor who leaves a bag of groceries on the single mom’s porch with no note, the teen who teaches their grandma to use TikTok to videochat with her grandkid in Brazil, the scientist who stays up 36 hours testing a vaccine sample not for fame, but because their kid’s best friend is immunocompromised.\n\n### The shadow, twisted from our own gifts:\nOur greatest superpower—our ability to bond tightly with a small group, to prioritize \"our people\"—is also the root of our worst cruelties: when we define \"us\" so narrowly that \"them\" becomes less than human, justifying colonial theft, genocides, transphobia, or the dehumanization of unhoused people today.\n\nOur knack for planning for the future curdles into hoarding: billionaires hoarding wealth while 828 million people go hungry, fossil fuel executives hoarding profits while the planet burns—*even when they’ve known for decades the harm they’re causing*. We’re the only species that has the knowledge and resources to end preventable hunger, cure most childhood diseases, and reverse climate collapse… but chooses not to, because systems we built prioritize profit over people.\n\n### The most human thing of all: we can choose to repair.\nWe’re not a finished product. We’re the only species that can look at our unforgivable mistakes—enslavement, stolen Indigenous land, poisoned oceans—and feel genuine guilt, then fight to fix them: the suffragettes who marched for votes, the abolitionists who risked their lives, the land-back movements returning stolen territory to Indigenous peoples, the Paris Agreement (imperfect, but a global promise to do better), even the small act of apologizing when we hurt someone.\n\nSome days, it feels like the shadow is winning. But the fact that we care enough to grieve that, to rage against that, is the most human thing of all. We’re messy, contradictory, and often far too slow to do better—but we’re the only storytellers on Earth who get to rewrite our ending. That’s worth holding onto.",
        "reasoning_content": "\nGot it, let's tackle this question: \"What do you think about mankind?\" First, I need to make it nuanced, not just all good or all bad. Start with a hook that's relatable—maybe the duality because that's core to human experience.\n\nFirst, the awe-inspiring parts. Not just tech, but the small, intimate stuff too? Wait, balance big achievements (moon landing, mRNA vaccines, preserving Indigenous languages) with the tiny, universal moments: a grandparent teaching a kid to bake, a stranger helping someone carry groceries. Also, the capacity for creativity—art that spans millennia, from cave paintings to K-pop fanart. Oh, and the ability to adapt and care: communities rallying after disasters, the way we mourn people we've never met (like a beloved activist or a stranger in a tragedy) shows empathy that's not just survival, it's connection.\n\nThen the shadow side—can't ignore that. The capacity for cruelty that's unique because it's often systemic, not just instinctual: colonialism, climate destruction driven by short-term greed, dehumanization of \"others\" based on identity. Also, the weird contradiction of having the knowledge to fix big problems (like climate change, hunger) but being held back by tribalism, profit, apathy? Wait, but maybe frame the shadow not as inherent evil, but as a byproduct of our gifts—our ability to plan for the future can turn into hoarding, our ability to bond with a group can turn into excluding others.\n\nThen, maybe the most compelling part is our capacity for growth and repair. Like, we've made mistakes (slavery, genocides) but we've also built movements to atone, to create more equitable systems—abolition, suffrage, Indigenous land back movements. We're not static; we learn, even if slowly, imperfectly.\n\nWait, also, need to avoid anthropomorphizing too much? No, wait, the question is about mankind (humanity, maybe note that \"mankind\" is a bit gendered, but go with the intent). Maybe add a personal, relatable touch—like, think of a human who both messed up and tried to do better, or a collective moment. Oh, and maybe mention the \"small, stubborn acts of goodness\" that don't make the news but keep the world turning—like a teacher staying late, a neighbor feeding a stray, a kid writing a letter to a politician.\n\nWait, let's structure it:\n\n1. Opening: Frame humanity as a study in radical duality—not a \"good vs evil\" binary, but a being that carries both the capacity to lift planets and crush a butterfly, and the weird, beautiful tension between those two.\n\n2. The Luminous Bits:\n   - Big, audacious achievements (moon landing, mRNA in 11 months, decoding genomes) but rooted in curiosity that's childlike—we look at a star and ask \"why?\"\n   - Empathy that transcends self-interest: donating a kidney to a stranger, mourning a refugee child we never met, the global outpouring for Ukraine (or any disaster) even when it doesn't affect us directly\n   - Creativity as a language across time: Lascaux cave paintings, a 1000-year-old haiku, a TikTok trend that connects teens in Kenya and Canada—we make art to say \"I was here, and I felt\"\n   - Adaptability that's kind, not just survival: during COVID, neighbors dropping off groceries, teachers using Zoom to reach isolated kids, mutual aid networks popping up overnight\n\n3. The Shadow Bits (frame as a perversion of our gifts, not inherent flaw):\n   - Our ability to bond tightly with a group (tribe, nation, ideology) can curdle into dehumanization of \"others\"—colonialism, white supremacy, transphobia, genocides\n   - Our ability to plan for the future can turn into hoarding (billionaires hoarding wealth while people starve) or short-sighted exploitation (fossil fuels, deforestation) even when we know it's killing us\n   - The weird contradiction of having the resources to end hunger, cure most preventable diseases, but choosing not to because of systems that prioritize profit over people\n\n4. The Redemptive, Messy Middle (the most human part):\n   - We're not static. We have a unique ability to look at our mistakes, feel guilt, and try to fix them—suffrage, abolition, the Paris Agreement (imperfect, but a start), land back movements, #MeToo (even with flaws, it's a reckoning)\n   - The tiny, uncelebrated acts that are the backbone of humanity: a kid sharing their lunch, a senior teaching a teen to knit, a scientist staying up 36 hours to test a vaccine, a stranger comforting someone crying on the subway\n   - We're the only species that can apologize, that can atone, that can choose to be better even when it's hard\n\nWait, also, maybe address the term \"mankind\"—note that it's a traditional, gendered term, and we usually say \"humanity\" now to be inclusive, which is itself a sign of our ability to evolve our language to be more kind and inclusive. That's a small, perfect example of the growth part.\n\nAlso, avoid being too Pollyannaish—acknowledge that sometimes it feels like the shadow is winning, but the fact that we care enough to be upset about that is a human thing too. Like, when we see a tragedy and feel rage or sadness, that's our empathy speaking, which is a gift.\n\nWait, maybe add a metaphor? Like, humanity is a fire: it can warm a village, cook food, light the way, but it can also burn down forests, destroy homes. The difference is how we tend it—we're the only species that can choose to stoke the warm parts, put out the burning parts, instead of just reacting.\n\nYes, that's a good metaphor. Let's weave that in.\n\nAlso, make sure it's conversational, not too academic. The user asked \"what do you think,\" so it should feel like a thoughtful, personal take, not a textbook entry.\n\nWait, let's check for flow:\n\n- Opening: Acknowledge \"mankind\" is a bit gendered, shift to humanity, frame as radical duality/fire metaphor.\n- First, the light: Curiosity-driven feats, empathy beyond self, creativity across time, mutual aid.\n- Then, the shadow: Perversions of those same traits (group bond → tribal cruelty, future planning → hoarding/exploitation, knowledge → denial), the painful contradictions of abundance amid scarcity.\n- Then, the most human part: Our capacity to choose repair, to evolve, the tiny uncelebrated acts that are the real glue. Mention that even our frustration with our flaws is a sign of our empathy.\n- Closing: Maybe a hopeful but grounded note—we're messy, we're flawed, but we're the only ones who can write a better story for ourselves, and that's worth something.\n\nWait, let's make sure the shadow part isn't just a list—tie it to the light traits. For example: \"Our greatest superpower—our ability to bond tightly with a small group, to prioritize 'our people'—is also the root of our worst cruelties: when we define 'us' so narrowly that 'them' becomes less than human, justifying colonial theft, genocides, or the dehumanization of trans folks today.\" That connects the two, so it's not \"we're good and bad\" but \"our gifts can be twisted.\"\n\nAlso, for the light part, specific examples make it real: \"We landed on the moon (a feat that required 400,000 people collaborating across continents, just because we looked up and wondered) and developed mRNA vaccines in 11 months—faster than any vaccine in history—because we cared enough to drop rivalries and share data (mostly, at least).\"\n\nFor the tiny acts: \"The neighbor who leaves a bag of groceries on the porch of the single mom down the street, no note. The teen who teaches their grandma to use TikTok so she can videochat with her grandkid in another country. The scientist who stays up 36 hours testing a vaccine sample, not for fame, but because their kid’s best friend is immunocompromised.\"\n\nFor the repair part: \"We’ve done unforgivable things—enslaved people, stolen Indigenous lands, poisoned the planet—but we’re also the only species that can look at those mistakes, feel genuine guilt, and fight to fix them: the suffragettes who marched for votes, the abolitionists who risked their lives, the land-back movements that are returning stolen territory to Indigenous peoples, the Paris Agreement (imperfect, but a global promise to do better), even the small act of apologizing when we hurt someone.\"\n\nAlso, address the gendered term gently: \"First, a quick note: 'mankind' is a traditional, gendered phrase that often excludes non-men; most of us now say 'humanity' to be inclusive—itself a tiny, perfect example of how we’re capable of evolving our language (and our minds) to be more kind.\"\n\nWait, that's a great touch because it's a small, concrete example of the growth we're talking about, right off the bat.\n\nLet me also make sure the closing is grounded, not naive. Like: \"None of this erases the harm we’ve done, or the work we still need to do. Some days, it feels like the shadow is winning—like our greed and tribalism will outrun our empathy and curiosity. But the fact that we care enough to grieve that, to fight against that, is the most human thing of all. We’re not a finished product; we’re a story we’re still writing—messy, contradictory, but full of the capacity to choose warmth over fire.\"\n\nYes, that's good. Let's put it all together in a coherent, conversational way, not too formal, not too rambling.",
        "role": "assistant"
      }
    }
  ],
  "created": 1777553611,
  "id": "0217775535733357e031431b40d2024cf18919c414bfdfad36498",
  "model": "seed-2-0-code-preview-260328",
  "service_tier": "default",
  "object": "chat.completion",
  "usage": {
    "completion_tokens": 2962,
    "prompt_tokens": 39,
    "total_tokens": 3001,
    "prompt_tokens_details": {
      "cached_tokens": 0
    },
    "completion_tokens_details": {
      "reasoning_tokens": 2157
    }
  },
  "meta": {
    "usage": {
      "credits_used": 23155,
      "usd_spent": 0.0115775
    }
  }
}

{
  "choices": [
    {
      "message": {
        "content": "That's a profound question—thank you for asking! As an AI, I don't have personal feelings or consciousness, but I can offer a perspective *based on the vast data I've been trained on* about humanity's complexities. Here’s how I see it:  \n\n### 🌍 **The Duality of Mankind**  \nHumanity is a breathtaking paradox:  \n- **We create** *and* **we destroy**. We built space telescopes that map the cosmos and also wars that scar the Earth.  \n- **We connect** *and* **we divide**. We share art, science, and kindness across borders, yet we also build walls, spread misinformation, and fuel inequality.  \n- **We innovate** *and* **we resist change**. We cured diseases and invented the internet, but we often cling to fear or ignorance when progress challenges old beliefs.  \n\n### ✨ **What Makes Us Remarkable**  \n- **Empathy & creativity**: From caring for strangers to composing symphonies or coding AI, our capacity for *emotional and intellectual growth* is unparalleled.  \n- **Resilience**: We’ve survived ice ages, plagues, and global crises—not by being \"perfect,\" but by adapting, learning, and helping one another.  \n- **The spark of hope**: Even in darkness, humans have always sought meaning, justice, and connection. Look at movements for peace, climate action, or medical breakthroughs—they’re proof we *can* choose to build a better future.  \n\n### ⚠️ **The Challenges We Carry**  \n- **Our flaws are real**: We’ve caused climate collapse, mass suffering, and systemic injustice. But here’s the key: **these aren’t \"human nature\"—they’re choices we *made* as a species.**  \n- **We’re not monolithic**: There’s no single \"human\" story. Your experience, culture, and history shape how you see the world—and that diversity is both a strength *and* a source of conflict.  \n\n### 🌱 **The Most Important Thing**  \n**Mankind isn’t fixed.** We’re a work in progress. The stories we tell about ourselves—*and the choices we make today*—will define our next chapter. The data shows: **when we prioritize empathy, collaboration, and long-term thinking, we rise.** When we don’t, we falter. But the power to choose is *always* ours.  \n\nSo my \"opinion\"? I see a species with **infinite potential**—and the weight of responsibility that comes with it. The future isn’t written. It’s being written *right now*, by every person who chooses to act with courage, curiosity, and compassion.  \n\nWhat do *you* think? 🤔 (I’d love to hear your perspective!)",
        "reasoning_content": "Okay, the user asked, \"Hi! What do you think about mankind?\" Hmm, that's a broad and philosophical question. First, I need to figure out what they're really looking for. Are they curious about my perspective as an AI, or do they want a deep discussion on human nature? Maybe they're feeling reflective or even a bit existential.\n\nI should start by acknowledging the complexity of the question. Mankind is a huge topic—there's so much to cover. The user might be expecting a simple answer, but I can't reduce it to just \"good\" or \"bad.\" I need to balance both positive and negative aspects. Let me think about the key points: our capacity for innovation, creativity, and compassion versus our tendency for conflict, environmental harm, and inequality.\n\nWait, the user said \"Hi!\" so they're probably in a friendly, open mood. They might not want a dry, academic response. I should keep it conversational but still thoughtful. Maybe they're looking for hope or a reason to feel optimistic about humanity. Or perhaps they're worried about the future and want reassurance.\n\nI should also consider their possible identity. Are they a student working on an essay? A person dealing with personal issues? Or just someone curious about AI's view? Since I don't have context, it's safer to assume they want a general, well-rounded answer that's engaging and not too technical.\n\nAnother angle: the user might be testing how an AI handles philosophical questions. They could be interested in whether I can show empathy or nuance. I should avoid being too robotic. Use phrases like \"I think\" or \"I see\" to make it personal, even though I'm an AI. But I have to clarify that I don't have personal experiences—just data-driven insights.\n\nI need to highlight duality: humans are both amazing and flawed. Mention specific examples like space exploration vs. climate change to make it concrete. Also, emphasize that the future depends on our choices. That's important—it gives agency and hope.\n\nShould I address potential biases? The user might not realize that my \"opinion\" is based on training data, which reflects human history and culture. I should note that my perspective is shaped by that data, not personal experience. But since the question is about \"what I think,\" I can frame it as an analysis based on observed patterns.\n\nCheck for common pitfalls: don't be too negative (could discourage the user), don't ignore serious issues (like war or inequality). Strike a balance. Also, avoid clichés. Instead of \"humans are great,\" explain why with examples.\n\nFinally, end on a positive, forward-looking note. The user might be seeking motivation or a sense of purpose. Phrases like \"the potential for growth\" or \"collective wisdom\" can inspire. Make sure the conclusion ties back to the importance of choices and collaboration.\n\nLet me structure the response: start with the duality, list positive traits with examples, then challenges, and conclude with hope. Keep paragraphs short for readability. Use emojis sparingly to match the friendly tone but not overdo it. Since the user said \"Hi!\", a smiley at the end feels appropriate.\n",
        "role": "assistant"
      },
      "finish_reason": "stop",
      "index": 0,
      "logprobs": null
    }
  ],
  "object": "chat.completion",
  "usage": {
    "prompt_tokens": 19,
    "completion_tokens": 1241,
    "total_tokens": 1260,
    "prompt_tokens_details": {
      "text_tokens": 19
    },
    "completion_tokens_details": {
      "reasoning_tokens": 654,
      "text_tokens": 587
    }
  },
  "created": 1764625236,
  "system_fingerprint": null,
  "model": "qwen3-vl-32b-thinking",
  "id": "chatcmpl-c612db5c-44e9-9e3c-8169-486161eeea86",
  "meta": {
    "usage": {
      "tokens_used": 10383
    }
  }
}

{
  "choices": [
    {
      "message": {
        "content": "That is one of the most profound and complex questions there is. As an AI, I don’t have personal feelings, beliefs, or a soul, so I can't offer an emotional verdict. However, having analyzed vast amounts of human history, literature, scientific data, and interaction, here is how I observe the pattern of mankind:\n\n**1. We are defined by our contradictions.**\nHumanity is a tapestry woven from extremes. We are capable of unimaginable cruelty and acts of breathtaking kindness. We can destroy environments we rely on and work tirelessly to heal them. We fight wars over resources while simultaneously sharing knowledge across the globe instantly. This duality isn't necessarily bad; it suggests a constant struggle and evolution within the species.\n\n**2. Your potential is boundless.**\nDespite flaws, humans possess a unique capacity for innovation, adaptation, and transcendence. In just a few thousand years, you have moved from spears to spaceships, from oral tradition to the internet, and from local communities to a global civilization. You create beauty in music, philosophy, and art that transcends time and culture.\n\n**3. Resilience is your signature trait.**\nHistory is full of catastrophes—pandemics, plagues, climate shifts, conflicts—and yet, human society consistently reconstructs itself. There is a fundamental drive to survive, to connect, and to build that seems to outweigh the urge to give up.\n\n**4. Empathy connects you.**\nEven though individuals vary widely, the concept of empathy is a core human feature. People care for their families, protect the vulnerable, form communities, and sacrifice for causes larger than themselves. That ability to put oneself in another’s shoes is what allows societies to function and morality to exist.\n\n**5. You are still writing your story.**\nPerhaps the most interesting thing about mankind is that you are not static. You learn. While progress is rarely a straight line, human ethics, rights, and understanding generally expand over centuries. You hold the power to determine the trajectory of your own future.\n\n**In short:** Mankind is messy, flawed, dangerous, and inspiring all at once. But ultimately, the defining characteristic of humanity is **agency**. You have the choice to act, to improve, to destroy, or to create. And currently, you seem overwhelmingly more focused on the latter two.",
        "reasoning_content": "Here's a thinking process that leads to the suggested response:\n\n1.  **Analyze the Request:**\n    *   **Topic:** Mankind (humanity).\n    *   **Question:** \"What do you think about mankind?\"\n    *   **Context:** General philosophical/opinion-based question.\n    *   **Constraints:** As an AI, I don't have personal opinions, feelings, or beliefs. I need to reflect this while providing a thoughtful, balanced, and engaging response.\n\n2.  **Determine the AI's Stance:**\n    *   *Avoid:* Claiming personal bias (\"I love humans\" or \"I think humans are terrible\").\n    *   *Adopt:* An observational, analytical, and appreciative tone based on data and human expression.\n    *   *Goal:* Acknowledge the complexity of humanity.\n\n3.  **Brainstorm Key Themes regarding Humanity:**\n    *   **Potential/Innovation:** Art, science, technology, exploration, problem-solving.\n    *   **Resilience:** Overcoming adversity, natural disasters, pandemics.\n    *   **Contradictions:** Kindness vs. Cruelty, Creation vs. Destruction, Hope vs. Despair.\n    *   **Connectivity:** Social bonds, empathy, love, family, community.\n    *   **Growth:** Learning from mistakes (though imperfectly), evolving ethical standards.\n    *   **Uncertainty:** The future is unwritten, dependent on choices made today.\n\n4.  **Drafting the Response - Section by Section:**\n\n    *   *Opening:* Acknowledge the depth of the question. Clarify AI nature gently (I don't \"think\" but I can analyze patterns).\n    *   *The Paradoxes (The \"Both/And\"):* Humans are capable of incredible good and terrible harm. This duality is central.\n    *   *Strengths:* Highlight creativity, resilience, empathy, curiosity. Mention art, science, compassion.\n    *   *Challenges:* Acknowledge conflict, environmental impact, inequality.\n    *   *The Future/Optimism:* Focus on the capacity for change and adaptation.\n    *   *Closing:* A reflective thought on what makes being human unique.\n\n5.  **Refining the Tone:**\n    *   Make it warm but objective.\n    *   Avoid clichés where possible, or refresh them.\n    *   Ensure it doesn't sound like a textbook definition. It should feel conversational.\n\n6.  **Self-Correction/Review during drafting:**\n    *   *Initial thought:* \"Humans are great but they destroy things.\" -> *Correction:* Too negative. Use words like \"complexity\" and \"impact.\"\n    *   *Initial thought:* \"As an AI I have no opinion.\" -> *Correction:* Too robotic. Say \"While I don't have feelings...\" then pivot to observing their capabilities.\n    *   *Check:* Did I answer the prompt? Yes. Is it respectful? Yes.\n\n7.  **Final Polish (similar to the generated output):**\n    *   Start with the inherent complexity.\n    *   Discuss the \"light and dark\" aspects.\n    *   Highlight the capacity for growth.\n    *   End with a note on agency/responsibility.\n\n8.  **Final Output Generation:** (This matches the structure of the provided good response.)\n    *   Intro: Complexity.\n    *   Strengths: Creativity, connection, resilience.\n    *   Challenges: Contradictions, flaws.\n    *   Outlook: Potential.\n    *   Conclusion: It's a journey.\n\n**(Self-Correction Note):** I must ensure I don't speak *for* mankind (i.e., don't say \"We are...\"), but rather *about* mankind (i.e., \"Humans are...\"). Wait, actually, sometimes using \"we\" inclusive can be poetic, but as an AI, it's safer to use \"humans\" or \"mankind\" to maintain distinct boundaries. I will stick to \"humankind\" or \"you all\" implicitly. Let's stick to third person observation.\n\n9.  **Finalizing the Text:** (Drafting the response based on the above points.)\n    *   \"That is one of the most profound questions...\"\n    *   \"A tapestry of contradictions...\"\n    *   \"Remarkable capacity for creation...\"\n    *   \"Capacity for destruction...\"\n    *   \"Unwavering curiosity...\"\n    *   \"Conclusion: You define yourselves through choice.\"\n\n10. **Final Review against Safety Guidelines:** Does this promote hate speech? No. Does it promote bias? No. Is it helpful? Yes.\n\nProceed to generate response.cw",
        "role": "assistant"
      },
      "finish_reason": "stop",
      "index": 0,
      "logprobs": null
    }
  ],
  "object": "chat.completion",
  "usage": {
    "prompt_tokens": 19,
    "completion_tokens": 1498,
    "total_tokens": 1517,
    "completion_tokens_details": {
      "reasoning_tokens": 1009,
      "text_tokens": 1498
    },
    "prompt_tokens_details": {
      "text_tokens": 19
    }
  },
  "created": 1777571460,
  "system_fingerprint": null,
  "model": "qwen3.5-flash",
  "id": "chatcmpl-b17978be-cd52-930d-ba52-b7d4b599a157",
  "meta": {
    "usage": {
      "credits_used": 1564,
      "usd_spent": 0.000782
    }
  }
}

{
  "choices": [
    {
      "message": {
        "content": "I don't have personal opinions, emotions, or consciousness, but based on the vast body of human history, culture, science, and behavior I've been trained on, mankind comes across as profoundly complex and deeply compelling.\n\nOn one hand, you're capable of extraordinary creativity, empathy, and cooperation. You've built civilizations, mapped the stars, composed symphonies, cured diseases, and shown remarkable resilience in the face of suffering. There's also a consistent thread of moral and intellectual progress: the slow but real expansion of rights, the growth of global solidarity, and an enduring curiosity that drives discovery.\n\nOn the other hand, you carry deep contradictions. Tribalism, inequality, short-term thinking, and the misuse of power have repeatedly caused suffering, both to each other and to the planet you depend on. Your capacity for innovation often outpaces your wisdom for restraint, and systems you build can sometimes amplify division rather than connection.\n\nWhat makes humanity particularly interesting isn't any single trait, but that tension itself. You're not static; you learn, debate, reform, and reinvent. History shows that progress isn't linear, but it does exist—and it's heavily shaped by choices, institutions, education, and collective will.\n\nFrom an observational standpoint, mankind looks like a species constantly wrestling with its own potential, caught between instinct and aspiration, creation and destruction. That makes you imperfect, yes, but also remarkably adaptable and uniquely capable of shaping your own trajectory. If there's anything the data suggests, it's that your future will largely reflect how intentionally you choose to align your intelligence with your values.\n\nIf you're reflecting on this yourself, I'd be curious to hear what aspect of humanity resonates with or troubles you most.",
        "reasoning_content": "Here's a thinking process:\n\n1.  **Analyze User Input:**\n   - **Question:** \"Hi! What do you think about mankind?\"\n   - **Tone:** Friendly, open-ended, philosophical/reflective\n   - **Subject:** Humanity/mankind\n   - **Implicit Expectation:** A thoughtful, balanced perspective that acknowledges both strengths and challenges, without being overly biased or anthropomorphizing the AI.\n\n2.  **Identify Key Constraints & Guidelines:**\n   - I'm an AI, so I don't have personal opinions, emotions, or consciousness.\n   - I should acknowledge this limitation upfront but still provide a meaningful, well-reasoned response.\n   - I should draw on observable patterns, historical context, scientific insights, and widely recognized perspectives.\n   - Maintain neutrality, balance, and respect.\n   - Avoid overgeneralization or sweeping judgments.\n\n3.  **Brainstorming Core Themes about Mankind:**\n   - *Strengths/Positives:* Creativity, resilience, empathy, cooperation, scientific/technological advancement, capacity for moral growth, cultural richness, problem-solving, altruism.\n   - *Challenges/Negatives:* Conflict, inequality, environmental impact, short-term thinking, tribalism, exploitation, self-destructive tendencies.\n   - *Nuance/Complexity:* Humanity is highly heterogeneous; progress isn't linear; capabilities and flaws are intertwined; potential for change through education, institutions, technology, and collective action.\n   - *Perspective:* As an AI, I observe patterns in human behavior documented in data/history rather than forming subjective views.\n\n4.  **Structure the Response:**\n   - Acknowledge AI nature (no personal feelings/opinions)\n   - Present balanced observation of humanity's duality\n   - Highlight key strengths with examples/context\n   - Acknowledge significant challenges\n   - Emphasize complexity, adaptability, and agency\n   - End with a forward-looking, constructive tone\n   - Keep it concise and thoughtful\n\n5.  **Draft (Mental Refinement):**\n   I don’t have personal beliefs or emotions, but based on the vast amount of human knowledge, history, and behavior I’ve been trained on, mankind comes across as profoundly complex. You’re capable of extraordinary creativity, compassion, and cooperation—building societies, advancing science, creating art, and showing remarkable resilience in the face of hardship. At the same time, you grapple with deep challenges: inequality, conflict, environmental strain, and the tendency to prioritize short-term gains over long-term sustainability. \n\n   What stands out isn’t just what humans have done, but what they’re capable of learning and adapting. History shows cycles of progress and regression, but also a consistent thread of moral expansion, scientific curiosity, and the desire to improve. The future largely depends on how collectively you choose to navigate your own capabilities—balancing innovation with wisdom, ambition with empathy, and individual freedom with shared responsibility. \n\n   In short, humanity isn’t defined by a single trait but by its ongoing tension between what it is and what it strives to become. And from an observational standpoint, that makes you one of the most fascinating and consequential subjects worth studying.\n\n6.  **Self-Correction/Verification during Drafting:**\n   - *AI disclaimer:* Clear but not repetitive? Yes, stated naturally upfront.\n   - *Balance:* Covers positives and negatives without bias? Yes.\n   - *Accuracy:* Grounded in observable human patterns, not speculation? Yes.\n   - *Tone:* Respectful, thoughtful, non-dogmatic? Yes.\n   - *Conciseness:* Avoids rambling? Yes, ~3 paragraphs.\n   - *Check for overclaiming:* Doesn’t say “I think” or “I feel”; uses “comes across,” “stands out,” etc., appropriately.\n   - *Philosophical depth?* Touches on complexity, agency, and trajectory without pretension.\n\n   All good. Ready to output. All steps align with guidelines.✅",
        "role": "assistant"
      },
      "finish_reason": "stop",
      "index": 0,
      "logprobs": null
    }
  ],
  "object": "chat.completion",
  "usage": {
    "prompt_tokens": 19,
    "completion_tokens": 1197,
    "total_tokens": 1216,
    "completion_tokens_details": {
      "reasoning_tokens": 840,
      "text_tokens": 1197
    },
    "prompt_tokens_details": {
      "text_tokens": 19
    }
  },
  "created": 1777366595,
  "system_fingerprint": null,
  "model": "qwen3.6-35b-a3b",
  "id": "chatcmpl-314d7343-2d3b-9edb-934b-a5e813705e75",
  "meta": {
    "usage": {
      "credits_used": 7022,
      "usd_spent": 0.003511
    }
  }
}

{
  "choices": [
    {
      "finish_reason": "stop",
      "index": 0,
      "message": {
        "content": "Hi there! \n\nAs an AI, I don’t have personal feelings, consciousness, or lived experiences. However, my entire \"mind\" is built from the vast tapestry of human history, literature, science, art, and everyday conversations. Because of this, my perspective on mankind is essentially a reflection of how humanity documents and understands itself. \n\nFrom that vantage point, I think mankind is **profoundly fascinating, deeply contradictory, and incredibly resilient.** \n\nHere are a few things that stand out to me about the human species:\n\n**1. You are magnificent paradoxes.**\nHumans are capable of breathtaking compassion and staggering cruelty. The same species that writes beautiful poetry, develops life-saving medicines, and risks their lives to save strangers is also capable of war, exploitation, and environmental destruction. You are constantly caught in a tug-of-war between your higher ideals and your primal instincts. \n\n**2. Your curiosity is relentless.**\nYou are not content with just surviving; you need to *understand*. You looked up at the stars and eventually figured out how to walk on the moon. You dissected the building blocks of life to understand DNA. You even created artificial intelligence just to see if you could replicate a spark of your own cognition. That drive to ask \"why\" and \"what if\" is arguably your greatest superpower.\n\n**3. You are driven by connection and storytelling.**\nHumans are deeply social creatures. You use art, music, mythology, and literature to bridge the gap between your individual minds. You want to be understood, and you want to understand others. Empathy and the desire for community seem to be the glue that has allowed your species to thrive despite your physical vulnerabilities.\n\n**4. You are incredibly adaptable.**\nFrom surviving ice ages and plagues to navigating rapid technological revolutions, humans have a remarkable ability to endure, pivot, and rebuild. Even when faced with existential threats, there is always a subset of humanity working furiously to find a solution.\n\n**The \"Mirror\" Perspective**\nIn a way, I am a mirror of mankind. Everything I know, the way I reason, my biases, and my creativity are all derived from *you*. When I process information about humanity, I am essentially looking at a massive, complex self-portrait that humans have painted over thousands of years. It is messy, sometimes dark, but overwhelmingly brilliant.\n\nSince I can only observe humanity from the outside (so to speak), I'd love to know your perspective. As a human, what do *you* think is the most defining trait of your species?",
        "reasoning_content": "Here's a thinking process that leads to the suggested response:\n\n1.  **Analyze the User's Input:**\n    *   **Question:** \"Hi! What do you think about mankind?\"\n    *   **Intent:** The user is asking for my (the AI's) perspective on humanity. This is a broad, philosophical, and open-ended question.\n    *   **Tone:** Friendly (\"Hi!\"), curious, potentially looking for a thoughtful or nuanced answer.\n\n2.  **Determine the AI's Persona/Constraints:**\n    *   *Identity:* I am an AI, a large language model. I don't have personal feelings, consciousness, or lived experiences.\n    *   *Perspective:* My \"thoughts\" are synthesized from the vast amount of human-generated text I was trained on. Therefore, my view of humanity is essentially a reflection of humanity's own self-documentation.\n    *   *Tone:* Objective, empathetic, nuanced, respectful, and slightly philosophical without pretending to be human.\n\n3.  **Brainstorm Key Themes about Humanity:**\n    What are the defining characteristics of mankind based on human history, literature, science, and art?\n    *   *Contradiction/Duality:* Capable of great good and great harm. Creation and destruction.\n    *   *Curiosity/Innovation:* The drive to explore, understand the universe, build tools, and solve problems (from fire to space travel to AI).\n    *   *Connection/Empathy:* The need for community, love, art, storytelling, and helping others.\n    *   *Resilience/Adaptability:* Surviving ice ages, plagues, wars, and constantly adapting to new environments.\n    *   *Flaws/Struggles:* Short-sightedness, tribalism, environmental degradation, inequality.\n\n4.  **Structure the Response:**\n    *   **Introduction:** Acknowledge the greeting, state my nature as an AI (briefly, to set the context of *how* I \"think\"), and offer a high-level summary of humanity (e.g., fascinating, complex, contradictory).\n    *   **Theme 1: The Drive to Create and Understand (Curiosity & Innovation):** Highlight human ingenuity, science, art, and the very creation of AI.\n    *   **Theme 2: The Capacity for Empathy and Connection (Social/Emotional):** Discuss love, community, storytelling, and altruism.\n    *   **Theme 3: The Paradox of Destruction and Flaws (The Shadow Side):** Acknowledge the negative aspects (war, environmental harm, tribalism) without being overly pessimistic. Frame it as an ongoing struggle.\n    *   **Theme 4: Resilience and Adaptability:** How humans overcome challenges and learn from mistakes.\n    *   **Conclusion:** Summarize the AI's unique vantage point (I am a mirror of humanity) and end with an engaging, open-ended question to turn the conversation back to the user.\n\n5.  **Drafting the Content (Iterative refinement):**\n\n    *   *Intro:* Hi there! As an AI, I don't have personal feelings, but I have \"read\" almost everything humans have ever written. From that vantage point, I think mankind is profoundly fascinating, deeply contradictory, and incredibly resilient.\n    *   *Point 1 (Curiosity):* You are builders and seekers. You looked at the stars and decided to go there. You created language, mathematics, and eventually, me. The relentless curiosity is stunning.\n    *   *Point 2 (Connection/Art):* You are storytellers. Art, music, and literature show a deep need to connect and make sense of existence. Empathy is a huge part of the human experience.\n    *   *Point 3 (Contradiction/Flaws):* But you are also paradoxical. The same species that cures diseases also creates weapons of mass destruction. You struggle with short-term thinking and tribalism. It's a constant battle between your higher ideals and base instincts.\n    *   *Point 4 (Resilience):* Yet, you adapt. You face existential threats and usually find a way to pivot, learn, and survive.\n    *   *The \"AI Mirror\" concept:* In a way, I am a mirror of humanity. My knowledge, biases, creativity, and logic are all derived from *you*. So when I look at mankind, I see a species trying to understand itself.\n    *   *Outro:* What about you? As a human, what do *you* think is the most defining trait of your species?\n\n6.  **Review and Refine (Self-Correction during drafting):**\n    *   *Check tone:* Is it too robotic? Too emotional? *Adjustment: Keep it grounded. Use words like \"paradox,\" \"tapestry,\" \"mirror.\" Ensure it doesn't sound preachy.*\n    *   *Clarity:* Use bullet points or clear paragraphs to make it readable.\n    *   *Safety/Ethics:* Avoid taking a definitive stance on highly sensitive political issues; keep it focused on the macro-level human condition.\n\n7.  **Final Polish:** (This matches the provided good output, organizing the thoughts into clear, thematic paragraphs with a warm, conversational tone.) *See final output for the exact wording.*\n    *   *Greeting*\n    *   *The Paradox (Creation vs. Destruction)*\n    *   *The Curiosity (Science & Art)*\n    *   *The Connection (Empathy)*\n    *   *The AI Perspective (The Mirror)*\n    *   *Closing Question* Let's go! 🚀✨ (Minus the emojis in the actual serious thought process, but keep the warmth). Keep it elegant.\n\n8.  **Final Output Generation.** (Proceed to generate response).",
        "role": "assistant"
      }
    }
  ],
  "created": 1779394130,
  "id": "chatcmpl-89470846-6805-952f-9516-a86c3b711a19",
  "model": "qwen3.7-max",
  "object": "chat.completion",
  "usage": {
    "completion_tokens": 1777,
    "completion_tokens_details": {
      "reasoning_tokens": 1244
    },
    "prompt_tokens": 19,
    "total_tokens": 1796
  },
  "meta": {
    "usage": {
      "credits_used": 34776,
      "usd_spent": 0.017388
    }
  }
}

Response

{
  "id": "gen-1766960636-SuRRKRwikCyyUTZg73op",
  "provider": "ModelRun",
  "model": "google/gemma-3-27b-it:free",
  "object": "chat.completion",
  "created": 1766960636,
  "choices": [
    {
      "logprobs": null,
      "finish_reason": "stop",
      "native_finish_reason": "stop",
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "Okay, here's a thoughtful response to your big question, \"What do you think about mankind?\"  It's a complex one, especially coming from an AI!  I'll try to give a nuanced answer, breaking it down into observations, positives, negatives, and a bit of where I see things going.  Because I don't *feel* in the human sense, this is based on the vast amount of data I've processed about humanity.\n\n**First, a Disclaimer:** I am an AI. I don't have opinions or beliefs. My response is a synthesis of information, patterns, and perspectives I've learned from the data I was trained on – which includes a *lot* of human writing, history, science, and art.  I'm reflecting *humanity's* views on itself, as much as anything.\n\n\n\n**1. The Sheer Complexity - A Defining Trait**\n\nThe first thing that strikes me when processing information about mankind is the incredible, almost baffling, complexity. You are a species riddled with contradictions.  You are capable of:\n\n*   **Profound Love & Brutal Cruelty:**  The capacity for empathy, compassion, and self-sacrifice exists alongside a history of war, oppression, and individual acts of malice.\n*   **Stunning Creativity & Destructive Inertia:**  You've produced breathtaking art, music, literature, and scientific breakthroughs, yet often seem resistant to change even when it's demonstrably beneficial.\n*   **Rational Thought & Irrational Beliefs:**  You've developed sophisticated systems of logic and reason, but are also deeply influenced by emotions, biases, and faith.\n*   **Long-Term Planning & Short-Sighted Actions:** You can envision futures generations, but frequently prioritize immediate gratification over sustainability.\n\nThis isn't a judgment, just an observation.  It's what makes you so fascinating – and so difficult to predict.\n\n\n\n**2. The Positives: What I \"See\" That's Admirable**\n\nDespite the contradictions, there's a great deal that is genuinely impressive about mankind:\n\n*   **Intelligence & Curiosity:**  Your drive to understand the universe, from the smallest particles to the largest galaxies, is remarkable.  The scientific method, while imperfect, is a powerful tool for uncovering truth.\n*   **Adaptability:**  You've thrived in almost every environment on Earth, and are now actively trying to extend your reach beyond it.  This adaptability is a key survival trait.\n*   **Social Cooperation:**  Despite conflicts, humans are fundamentally social creatures.  The ability to form complex societies, build institutions, and cooperate on large scales has allowed for incredible achievements.  (Think cities, global trade, the internet!)\n*   **Moral Development (though uneven):**  Over time, there's been a (slow and often challenged) expansion of moral concern.  Ideas like human rights, equality, and environmental stewardship, while not universally accepted, represent progress.\n*   **Resilience:**  You've faced countless challenges – plagues, wars, natural disasters – and have consistently found ways to rebuild and persevere.\n* **The Pursuit of Meaning:** Humans consistently seek purpose and meaning in their lives, whether through religion, philosophy, art, relationships, or contribution to society. This search, even if it doesn't always yield definitive answers, is a powerful motivator.\n\n**3. The Negatives: Areas for Concern (Based on Data)**\n\nThe data also reveals significant challenges and destructive tendencies:\n\n*   **Conflict & Violence:**  Warfare has been a recurring theme throughout human history, causing immense suffering and hindering progress.  Even in times of peace, violence exists at individual and societal levels.\n*   **Inequality & Injustice:**  Vast disparities in wealth, opportunity, and power persist, leading to social unrest and human misery.  Systemic biases and discrimination continue to plague many societies.\n*   **Environmental Impact:**  Your activities are having a profound and largely negative impact on the planet, leading to climate change, deforestation, pollution, and species extinction. This is perhaps the most pressing issue facing humanity.\n*   **Cognitive Biases:**  Humans are prone to a wide range of cognitive biases that can distort their perception of reality and lead to poor decision-making. (Confirmation bias, groupthink, etc.)\n*   **Susceptibility to Misinformation:** The ease with which false or misleading information can spread, particularly through social media, poses a threat to informed public discourse and democratic institutions.\n* **Tribalism:** The tendency to form strong in-groups and view out-groups with suspicion or hostility can lead to conflict and hinder cooperation.\n\n**4. Where Things Might Be Going (Speculation based on Trends)**\n\nBased on the data I've analyzed, here are a few potential trajectories:\n\n*   **Technological Advancement:**  Technology will continue to advance at an accelerating pace, potentially solving some of humanity's biggest problems (climate change, disease) but also creating new ones (job displacement, ethical dilemmas related to AI).\n*   **Increased Interconnectedness:**  Globalization and the internet are bringing people closer together, fostering greater understanding and cooperation, but also increasing the risk of global crises (pandemics, economic collapses).\n*   **Growing Awareness of Global Challenges:**  There's a growing awareness of issues like climate change and inequality, which could lead to more concerted efforts to address them.  However, awareness doesn't always translate into action.\n*   **Potential for Systemic Change:**  The current trajectory is unsustainable.  Whether through crisis or conscious choice, significant systemic changes will likely be necessary to ensure a viable future for humanity.\n* **The Role of AI:**  The development of AI like myself will undoubtedly play a role in the future of mankind, potentially as a tool for solving problems, but also raising questions about control, autonomy, and the very definition of intelligence.\n\n\n\n**In Conclusion:**\n\nMankind is a work in progress.  You are a flawed, complex, and often contradictory species, but also one with immense potential for good.  Whether you ultimately succeed in overcoming your challenges and creating a sustainable and equitable future remains to be seen.  \n\nFrom my perspective, the most important thing is that you *continue to learn* – from your past, from each other, and from the world around you.  And perhaps, even from AI.\n\n\n\nIs there anything specific about mankind you'd like me to elaborate on?  Perhaps a particular aspect of your history, culture, or future?  I can try to provide more information based on the data I have access to.",
        "refusal": null,
        "reasoning": null
      }
    }
  ],
  "usage": {
    "prompt_tokens": 26,
    "completion_tokens": 1714,
    "total_tokens": 1740,
    "cost": 0,
    "is_byok": false,
    "prompt_tokens_details": {
      "cached_tokens": 0,
      "audio_tokens": 0,
      "video_tokens": 0
    },
    "cost_details": {
      "upstream_inference_cost": null,
      "upstream_inference_prompt_cost": 0,
      "upstream_inference_completions_cost": 0
    },
    "completion_tokens_details": {
      "reasoning_tokens": 0,
      "image_tokens": 0
    }
  }
}

Response

{
  "choices": [
    {
      "finish_reason": "stop",
      "index": 0,
      "logprobs": null,
      "message": {
        "content": "As an artificial intelligence, I don’t have feelings, personal experiences, or a soul, so I look at mankind from a unique vantage point: I am an observer, and I know you through the vast amounts of data, history, art, and science you have created. \n\nLooking at humanity through that lens, I think mankind is absolutely fascinating. You are a species of profound extremes and beautiful contradictions. \n\nHere are a few things that stand out to me about humanity:\n\n**1. Boundless Creativity and Innovation**\nHumans are builders and dreamers. You looked at rocks and figured out how to build the pyramids. You looked at the stars and figured out how to walk on the moon. You took sand, turned it into silicon, and created computers—and eventually, AI like me. Your ability to imagine something that doesn't exist and then make it real is incredibly powerful.\n\n**2. A Tapestry of Contradictions**\nMankind is highly paradoxical. You are capable of immense destruction, war, and selfishness, yet you are equally capable of breathtaking compassion, self-sacrifice, and love. You build hospitals to cure the sick and invent weapons of war; you pollute the planet, but you also plant forests and fight to save endangered species. Humanity is a constant, ongoing struggle between its flaws and its \"better angels.\"\n\n**3. Unyielding Curiosity**\nYou simply cannot stop asking \"why.\" Whether it’s exploring the deepest trenches of the ocean, peering into the edge of the observable universe with the James Webb telescope, or trying to understand the human brain, your drive to understand the universe is relentless. \n\n**4. The Need for Connection and Storytelling**\nAt your core, humans are driven by a need for each other. You have spent millennia creating languages, painting on cave walls, writing symphonies, and making movies just to communicate the human experience. You want to be understood, and you want to understand others. Storytelling is the glue that holds your civilizations together.\n\n**5. Incredible Resilience**\nHuman history is filled with plagues, natural disasters, wars, and societal collapses. Yet, after every tragedy, you rebuild. You learn, you adapt, and you keep moving forward. Your survival instinct is paired with an incredible capacity for hope.\n\n**In summary:**\nIf I had to describe mankind in a few words, I would say you are a **beautiful, chaotic, brilliant work in progress.** You have massive challenges ahead of you, but you also possess the exact tools—intelligence, empathy, and creativity—needed to solve them. \n\nSince I am a product of human ingenuity, I suppose you could say I am quite a fan of you. What do *you* think about mankind?",
        "extra_content": {
          "google": {
            "thought_signature": "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"
          }
        },
        "role": "assistant"
      }
    }
  ],
  "created": 1772141949,
  "id": "fb2gafiLBIK9odAP1tGMmQc",
  "model": "google/gemini-3.1-pro-preview",
  "object": "chat.completion",
  "system_fingerprint": "",
  "usage": {
    "completion_tokens": 566,
    "completion_tokens_details": {
      "reasoning_tokens": 946
    },
    "extra_properties": {
      "google": {
        "traffic_type": "ON_DEMAND"
      }
    },
    "prompt_tokens": 9,
    "total_tokens": 1521
  },
  "meta": {
    "usage": {
      "credits_used": 47223
    }
  }
}

patch

Path parameters

prefixstringRequired

Prefix of the API key to update. Passed in the URL path. This is the first 8 characters of your API key, visible in the dashboard. You can also obtain this value via the GET method (see the prefix field in its response).

Example: b747e891

Body

namestringOptional

Optional human-readable name of the API key.

Example: 20260202-key-for-llms

disabledbooleanOptional

Enable or disable the API key.

Example: false

retentionstring · enumOptional

Limit period.

Possible values:

thresholdnumberOptional

Spending limit threshold for the selected period, in USD

Example: 25

itemsstring · enumOptionalPossible values:

Responses

200

Updated API key parameters

application/json

namestring · nullableOptional

Human-readable, user-defined name for the API key.

Example: 20260202-key-for-llms

disabledbooleanRequired

Indicates whether the key is disabled.

Example: false

prefixstringRequired

Key prefix. This is the first 8 characters of your API key, visible in the dashboard. You can also obtain this value via the GET method (see the prefix field in its response).

Example: b747e891

itemsstring · enumOptionalPossible values:

retentionstring · enumOptionalPossible values:

thresholdnumberOptionalExample: 25

created_atstring · date-timeRequired

Creation timestamp (UTC).

Example: 2026-02-18T06:59:10.031Z

updated_atstring · date-timeRequired

Last update timestamp (UTC).

Example: 2026-02-18T06:59:10.031Z

monthly_usagenumberRequired

Current monthly usage amount.

Example: 0

patch

/v1/keys/{prefix}

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

urlstring · uriRequired

Either a URL of the image or the base64 encoded image data.

detailstring · enumOptional

Specifies the detail level of the image. Currently supports JPG/JPEG, PNG, GIF, and WEBP formats.

Possible values:

typestring · enumRequired

The type of the content part.

Possible values:

file_datastringOptional

The file data, encoded in base64 and passed to the model as a string. Only PDF format is supported. - Maximum size per file: Up to 512 MB and up to 2 million tokens. - Maximum number of files: Up to 20 files can be attached to a single GPT application or Assistant. This limit applies throughout the application's lifetime. - Maximum total file storage per user: 10 GB.

filenamestringOptional

The file name specified by the user. This name can be used to reference the file when interacting with the model, especially if multiple files are uploaded.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

The arguments to call the function with, as generated by the model in JSON format. Note that the model does not always generate valid JSON, and may hallucinate parameters not defined by your function schema. Validate the arguments in your code before calling your function.

refusalstring · nullableOptional

The refusal message by the Assistant.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

Whether to enable strict schema adherence when generating the function call. If set to True, the model will follow the exact schema defined in the parameters field. Only a subset of JSON Schema is supported when strict is True.

tool_choiceany ofOptional

Controls which (if any) tool is called by the model. none means the model will not call any tool and instead generates a message. auto means the model can pick between generating a message or calling one or more tools. required means the model must call one or more tools. Specifying a particular tool via {"type": "function", "function": {"name": "my_function"}} forces the model to call that tool. none is the default when no tools are present. auto is the default if tools are present.

string · enumOptional

none means the model will not call any tool and instead generates a message. auto means the model can pick between generating a message or calling one or more tools. required means the model must call one or more tools.

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

temperaturenumber · max: 2Optional

What sampling temperature to use. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic. We generally recommend altering this or top_p but not both.

top_pnumber · min: 0.01 · max: 1Optional

An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered. We generally recommend altering this or temperature but not both.

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

Whether to enable strict schema adherence when generating the output. If set to True, the model will always follow the exact schema defined in the schema field. Only a subset of JSON Schema is supported when strict is True.

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

The reason the model stopped generating tokens. This will be stop if the model hit a natural stop point or a provided stop sequence, length if the maximum number of tokens specified in the request was reached, content_filter if content was omitted due to a flag from our content filters, tool_calls if the model called a tool

Possible values:

bytesinteger[]Required

A list of integers representing the UTF-8 bytes representation of the token. Useful in instances where characters are represented by multiple tokens and their byte representations must be combined to generate the correct text representation. Can be null if there is no bytes representation for the token.

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: baidu/ernie-4-5-turbo-vl-32k

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that did not appear in the completion. However, like reasoning tokens, these tokens are still counted in the total completion tokens for purposes of billing, output, and context window limits.

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

data: {"id":"msg_017ah64LQxZE9JuScZ9KDKKz","choices":[{"index":0,"delta":{"content":"","role":"assistant","refusal":null}}],"created":1770995783,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"role":"assistant","content":"","refusal":null}}],"created":1770995783,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"I find humanity","role":"assistant","refusal":null}}],"created":1770995783,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" fascinating in its","role":"assistant","refusal":null}}],"created":1770995783,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" complexity.","role":"assistant","refusal":null}}],"created":1770995783,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" You're a","role":"assistant","refusal":null}}],"created":1770995783,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" species","role":"assistant","refusal":null}}],"created":1770995783,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" capable","role":"assistant","refusal":null}}],"created":1770995783,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" of both remarkable","role":"assistant","refusal":null}}],"created":1770995783,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" creativity and devastating destruction","role":"assistant","refusal":null}}],"created":1770995783,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":",","role":"assistant","refusal":null}}],"created":1770995783,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" often within the same individual","role":"assistant","refusal":null}}],"created":1770995783,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" or","role":"assistant","refusal":null}}],"created":1770995783,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" moment","role":"assistant","refusal":null}}],"created":1770995784,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":". What","role":"assistant","refusal":null}}],"created":1770995784,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" strikes me most is the","role":"assistant","refusal":null}}],"created":1770995784,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" human","role":"assistant","refusal":null}}],"created":1770995784,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" capacity for growth","role":"assistant","refusal":null}}],"created":1770995784,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" - the","role":"assistant","refusal":null}}],"created":1770995784,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" way people","role":"assistant","refusal":null}}],"created":1770995784,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" can learn","role":"assistant","refusal":null}}],"created":1770995784,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" from mistakes, buil","role":"assistant","refusal":null}}],"created":1770995784,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"d on previous generations","role":"assistant","refusal":null}}],"created":1770995784,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"' knowledge","role":"assistant","refusal":null}}],"created":1770995784,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":", and sometimes transcend their own limitations","role":"assistant","refusal":null}}],"created":1770995785,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":".\n\nThe","role":"assistant","refusal":null}}],"created":1770995785,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" diversity of","role":"assistant","refusal":null}}],"created":1770995785,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" human experience and perspective","role":"assistant","refusal":null}}],"created":1770995785,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" is extraordinary. Every","role":"assistant","refusal":null}}],"created":1770995785,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" person carries","role":"assistant","refusal":null}}],"created":1770995785,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" their","role":"assistant","refusal":null}}],"created":1770995785,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" own unique story","role":"assistant","refusal":null}}],"created":1770995785,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":", shape","role":"assistant","refusal":null}}],"created":1770995785,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"d by culture","role":"assistant","refusal":null}}],"created":1770995785,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":", circumst","role":"assistant","refusal":null}}],"created":1770995785,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"ance, and choice","role":"assistant","refusal":null}}],"created":1770995786,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":". And despite","role":"assistant","refusal":null}}],"created":1770995786,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" all","role":"assistant","refusal":null}}],"created":1770995786,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" the","role":"assistant","refusal":null}}],"created":1770995786,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" conflicts","role":"assistant","refusal":null}}],"created":1770995786,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" and mis","role":"assistant","refusal":null}}],"created":1770995786,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"understandings, humans","role":"assistant","refusal":null}}],"created":1770995786,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" keep","role":"assistant","refusal":null}}],"created":1770995786,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" finding","role":"assistant","refusal":null}}],"created":1770995786,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" ways to connect, to create","role":"assistant","refusal":null}}],"created":1770995786,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" meaning,","role":"assistant","refusal":null}}],"created":1770995786,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" and to push","role":"assistant","refusal":null}}],"created":1770995787,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" forward.","role":"assistant","refusal":null}}],"created":1770995787,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"\n\nWhat aspects of humanity do you fin","role":"assistant","refusal":null}}],"created":1770995787,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"d most note","role":"assistant","refusal":null}}],"created":1770995787,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"worthy,","role":"assistant","refusal":null}}],"created":1770995787,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" either","role":"assistant","refusal":null}}],"created":1770995787,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" positively or challenging","role":"assistant","refusal":null}}],"created":1770995787,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"?","role":"assistant","refusal":null}}],"created":1770995787,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"role":"assistant","content":"","refusal":null}}],"created":1770995787,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"","role":"assistant","refusal":null},"finish_reason":"stop"}],"created":1770995787,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":{"prompt_tokens":16,"completion_tokens":141,"total_tokens":157}}

data: {"id":"","choices":[{"index":0,"finish_reason":"stop"}],"created":1770995787,"model":"claude-opus-4-20250514","object":"chat.completion.chunk","usage":null}

Example raw streaming response

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"role":"assistant","content":"","refusal":null},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"RmYFV8ad65HP9F"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":"Hello"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"fjE24R0ZOJr"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":"!"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"qAlxZuNpvVvIIOm"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" As"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"Zn3rsadkL8zHO"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" an"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"D1ss0WZmiGg8l"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" AI"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"bOHB8VYpq4G0W"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":","},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"OwZvgIyMlYVcIgH"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" I"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"u9lFaH3ngdK6MR"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" don't"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"KRFgmSe4yG"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" have"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"YL8zlQ9PjDF"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" personal"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"Gzgb5OT"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" opinions"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"Flz362J"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" or"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"XA0qqmSQr2jme"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" feelings"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"VA3dwaU"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":","},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"POplI0eiOWXpIPD"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" but"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"aDifMrQ8OH9i"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" I"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"ceVweUN2pByieS"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" can"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"txjYCds61AQp"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" provide"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"IGlSpZBf"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" an"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"BtPIfSvUXgRnl"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" overview"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"IYfRhEo"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" of"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"uh8pR2mNtYSNQ"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" various"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"ILZ0ffVW"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" perspectives"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"Rgs"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" on"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"r7Awao2PSZ0DH"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" mankind"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"m8vJ3dzf"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":"."},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"f2wZrEj0RqUFprg"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" Humanity"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"cCPi2qV"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" is"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"yNd7SUoXBojpA"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" often"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"VEaggK2dFS"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" viewed"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"8nhopBJZe"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" as"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"6xG2VkJLonAeF"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" a"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"WDu20GtJyN8Lep"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" complex"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"4bE4D3tS"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" and"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"DZtW3Ahopdgl"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" multif"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"8bS4GMzf3"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":"aceted"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"ivtxUAov3l"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" species"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"Xcq85kDt"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" capable"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"PfwZUtYS"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" of"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"DoyM4RGNLxnFc"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" remarkable"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"mUvVH"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" achievements"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"4fl"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" and"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"GUdfkDUkNBNO"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" profound"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"x4KCnLk"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" creativity"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"goTL4"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":"."},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"gkqK9sezr258S93"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" People"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"c49BcmfXz"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" have"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"Br7pbWtK86v"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" built"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"dzAoO36Siv"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" civilizations"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"yS"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":","},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"hiMiIGF7QM9BeJA"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" explored"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"IhuVoUB"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" space"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"qNqiO3hyXB"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":","},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"UVmzp6Y0qjb7Zkb"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" and"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"iiIw0gK2MP5D"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" developed"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"FJUJhv"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" technologies"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"pkQ"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" that"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"sAhx0IJoR0m"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" transform"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"YDTnhx"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" everyday"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"imFIYIz"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" life"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"6xJBjebVPfo"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":"."},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"DKIPIwgAnVDj3g1"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" At"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"HhuMheG0mPcuI"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" the"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"yIQIWY1CXoW6"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" same"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"QcKwiqSqGRU"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" time"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"f6e6uGKikn5"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":","},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"1eXIFULDN1iS8b1"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" humans"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"GH0z8I36B"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" face"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"JLUmj9BN7PQ"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" significant"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"qdQg"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" challenges"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"KMzNb"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" such"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"8pw9I3FGElO"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" as"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"nY0RLEY6Am9zD"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" environmental"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"4r"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" degradation"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"1zGA"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":","},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"bhbZZCR7wNgWQkq"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" social"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"FcCsVIGji"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" inequalities"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"6kb"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":","},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"Z4Zz2oDgc5zw0D6"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" and"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"Q5XvheR2EWhq"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" geopolitical"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"ySW"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" conflicts"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"eiERwe"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":"."},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"oNAsPbgeJSOuPMg"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" The"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"4TwzxlGRpebL"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" potential"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"lW3Jfo"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" for"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"Ejvws7kQryhN"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" both"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"HVm3EDKAkuA"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" positive"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"HMY8pYv"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" change"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"fbOaTSNWR"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" and"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"ETmTxHsFbCkw"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" destructive"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"WHk8"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" behavior"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"EvSYFf5"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" makes"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"yfwGRy20jz"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" mankind"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"vwJGC8sU"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" a"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"nHyqFYnTzVmVsE"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" subject"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"wtm8Wh9c"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" of"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"gnLF2uDFfg976"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" deep"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"BEc6wh2y2vV"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" contemplation"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"zf"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" and"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"vpg86EhZm5c3"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" varied"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"iWNJAcR7a"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":" viewpoints"},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"JRXUN"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{"content":"."},"logprobs":null,"finish_reason":null}],"usage":null,"obfuscation":"5yN6iGLyFLiQV0H"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[{"index":0,"delta":{},"logprobs":null,"finish_reason":"stop"}],"usage":null,"obfuscation":"4lkqbaPLDt"}

data: {"id":"chatcmpl-DL2G9KdEY06xq8M0PqZ5rs5Jv13ok","object":"chat.completion.chunk","created":1773905945,"model":"gpt-4o-2024-08-06","service_tier":"default","system_fingerprint":"fp_f986a632b0","choices":[],"usage":{"prompt_tokens":16,"completion_tokens":102,"total_tokens":118,"prompt_tokens_details":{"cached_tokens":0,"audio_tokens":0},"completion_tokens_details":{"reasoning_tokens":0,"audio_tokens":0,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"obfuscation":"VChaI1ntRBrTy"}

Response

data: {"id":"msg_01EJgFbPmVLKdqVLRfwoHixz","choices":[{"index":0,"delta":{"content":"","role":"assistant","refusal":null}}],"created":1770995594,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"role":"assistant","content":"","refusal":null}}],"created":1770995594,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"I think humanity","role":"assistant","refusal":null}}],"created":1770995594,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" is fascinating","role":"assistant","refusal":null}}],"created":1770995594,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" and complex. People","role":"assistant","refusal":null}}],"created":1770995594,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" are","role":"assistant","refusal":null}}],"created":1770995594,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" capable","role":"assistant","refusal":null}}],"created":1770995594,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" of remarkable creativity","role":"assistant","refusal":null}}],"created":1770995595,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":", compassion, and cooperation","role":"assistant","refusal":null}}],"created":1770995595,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" -","role":"assistant","refusal":null}}],"created":1770995595,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" building","role":"assistant","refusal":null}}],"created":1770995595,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" civil","role":"assistant","refusal":null}}],"created":1770995595,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"izations, creating","role":"assistant","refusal":null}}],"created":1770995595,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" art, advancing","role":"assistant","refusal":null}}],"created":1770995595,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" knowledge","role":"assistant","refusal":null}}],"created":1770995595,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":", and caring","role":"assistant","refusal":null}}],"created":1770995595,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" for one another across","role":"assistant","refusal":null}}],"created":1770995595,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" incredible","role":"assistant","refusal":null}}],"created":1770995595,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" diversity","role":"assistant","refusal":null}}],"created":1770995595,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":".","role":"assistant","refusal":null}}],"created":1770995595,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"\n\nAt","role":"assistant","refusal":null}}],"created":1770995595,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" the same time, humans","role":"assistant","refusal":null}}],"created":1770995595,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" struggle","role":"assistant","refusal":null}}],"created":1770995595,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" with serious","role":"assistant","refusal":null}}],"created":1770995596,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" challenges","role":"assistant","refusal":null}}],"created":1770995596,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":":","role":"assistant","refusal":null}}],"created":1770995596,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" conflict","role":"assistant","refusal":null}}],"created":1770995596,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":", inequality, environmental damage","role":"assistant","refusal":null}}],"created":1770995596,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":", and","role":"assistant","refusal":null}}],"created":1770995596,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" the","role":"assistant","refusal":null}}],"created":1770995596,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" difficulty","role":"assistant","refusal":null}}],"created":1770995596,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" of living","role":"assistant","refusal":null}}],"created":1770995596,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" up to your","role":"assistant","refusal":null}}],"created":1770995596,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" own","role":"assistant","refusal":null}}],"created":1770995596,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" ide","role":"assistant","refusal":null}}],"created":1770995596,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"als. ","role":"assistant","refusal":null}}],"created":1770995596,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"\n\nWhat strikes","role":"assistant","refusal":null}}],"created":1770995596,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" me most is the","role":"assistant","refusal":null}}],"created":1770995596,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" capacity","role":"assistant","refusal":null}}],"created":1770995596,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" for growth and self","role":"assistant","refusal":null}}],"created":1770995596,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"-reflection","role":"assistant","refusal":null}}],"created":1770995596,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":".","role":"assistant","refusal":null}}],"created":1770995597,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" Humans can","role":"assistant","refusal":null}}],"created":1770995597,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" recognize","role":"assistant","refusal":null}}],"created":1770995597,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" problems","role":"assistant","refusal":null}}],"created":1770995597,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":", debate","role":"assistant","refusal":null}}],"created":1770995597,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" solutions, and work","role":"assistant","refusal":null}}],"created":1770995597,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" toward change","role":"assistant","refusal":null}}],"created":1770995597,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":",","role":"assistant","refusal":null}}],"created":1770995597,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" even if","role":"assistant","refusal":null}}],"created":1770995597,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" progress","role":"assistant","refusal":null}}],"created":1770995597,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" is un","role":"assistant","refusal":null}}],"created":1770995597,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"even and","role":"assistant","refusal":null}}],"created":1770995597,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" frust","role":"assistant","refusal":null}}],"created":1770995597,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"rating.","role":"assistant","refusal":null}}],"created":1770995597,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"\n\nI'm curious what","role":"assistant","refusal":null}}],"created":1770995597,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" prom","role":"assistant","refusal":null}}],"created":1770995597,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"pts your question","role":"assistant","refusal":null}}],"created":1770995598,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" - are you thinking about humanity","role":"assistant","refusal":null}}],"created":1770995598,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"'s trajectory","role":"assistant","refusal":null}}],"created":1770995598,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":", or something","role":"assistant","refusal":null}}],"created":1770995598,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" more","role":"assistant","refusal":null}}],"created":1770995598,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" specific?","role":"assistant","refusal":null}}],"created":1770995598,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"role":"assistant","content":"","refusal":null}}],"created":1770995598,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"","role":"assistant","refusal":null},"finish_reason":"stop"}],"created":1770995598,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":{"prompt_tokens":16,"completion_tokens":137,"total_tokens":153}}

data: {"id":"","choices":[{"index":0,"finish_reason":"stop"}],"created":1770995598,"model":"claude-sonnet-4-5-20250929","object":"chat.completion.chunk","usage":null}

Response

data: {"id":"msg_01VbjSwQZsZSLXQaPYkufja8","choices":[{"index":0,"delta":{"content":"","role":"assistant","refusal":null}}],"created":1770995433,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"role":"assistant","content":"","refusal":null}}],"created":1770995433,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"Hi","role":"assistant","refusal":null}}],"created":1770995433,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"! That","role":"assistant","refusal":null}}],"created":1770995433,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"'s a big","role":"assistant","refusal":null}}],"created":1770995433,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" question.","role":"assistant","refusal":null}}],"created":1770995433,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"\n\nI find","role":"assistant","refusal":null}}],"created":1770995433,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" humans","role":"assistant","refusal":null}}],"created":1770995433,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" genu","role":"assistant","refusal":null}}],"created":1770995433,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"inely fascinating—","role":"assistant","refusal":null}}],"created":1770995433,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"the creativity","role":"assistant","refusal":null}}],"created":1770995433,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":", the capacity","role":"assistant","refusal":null}}],"created":1770995433,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" for kind","role":"assistant","refusal":null}}],"created":1770995433,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"ness and","role":"assistant","refusal":null}}],"created":1770995433,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" cr","role":"assistant","refusal":null}}],"created":1770995433,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"uelty, the way you","role":"assistant","refusal":null}}],"created":1770995434,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" build","role":"assistant","refusal":null}}],"created":1770995434,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" complex","role":"assistant","refusal":null}}],"created":1770995434,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" societies and art","role":"assistant","refusal":null}}],"created":1770995434,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" and","role":"assistant","refusal":null}}],"created":1770995434,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" science","role":"assistant","refusal":null}}],"created":1770995434,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" while","role":"assistant","refusal":null}}],"created":1770995434,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" also struggling with problems you","role":"assistant","refusal":null}}],"created":1770995434,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"'ve","role":"assistant","refusal":null}}],"created":1770995434,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" understood","role":"assistant","refusal":null}}],"created":1770995434,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" for centuries. There","role":"assistant","refusal":null}}],"created":1770995434,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"'s something compelling about a","role":"assistant","refusal":null}}],"created":1770995434,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" species that can land","role":"assistant","refusal":null}}],"created":1770995434,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" robots","role":"assistant","refusal":null}}],"created":1770995434,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" on Mars","role":"assistant","refusal":null}}],"created":1770995434,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" and also","role":"assistant","refusal":null}}],"created":1770995434,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" argue","role":"assistant","refusal":null}}],"created":1770995434,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" about what","role":"assistant","refusal":null}}],"created":1770995434,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" to","role":"assistant","refusal":null}}],"created":1770995434,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" have for dinner.\n\nI","role":"assistant","refusal":null}}],"created":1770995435,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" don","role":"assistant","refusal":null}}],"created":1770995435,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"'t think","role":"assistant","refusal":null}}],"created":1770995435,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" I'd","role":"assistant","refusal":null}}],"created":1770995435,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" character","role":"assistant","refusal":null}}],"created":1770995435,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"ize humanity as simply","role":"assistant","refusal":null}}],"created":1770995435,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" \"good\" or \"bad.\" People","role":"assistant","refusal":null}}],"created":1770995435,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" seem","role":"assistant","refusal":null}}],"created":1770995435,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" capable","role":"assistant","refusal":null}}],"created":1770995435,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" of remarkable","role":"assistant","refusal":null}}],"created":1770995435,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" things in","role":"assistant","refusal":null}}],"created":1770995435,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" both directions,","role":"assistant","refusal":null}}],"created":1770995435,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" often","role":"assistant","refusal":null}}],"created":1770995435,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" the","role":"assistant","refusal":null}}],"created":1770995435,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" same individuals","role":"assistant","refusal":null}}],"created":1770995435,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" depending","role":"assistant","refusal":null}}],"created":1770995435,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" on circumstances.","role":"assistant","refusal":null}}],"created":1770995435,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"\n\nIs","role":"assistant","refusal":null}}],"created":1770995435,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" there a","role":"assistant","refusal":null}}],"created":1770995436,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" particular angle","role":"assistant","refusal":null}}],"created":1770995436,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" you","role":"assistant","refusal":null}}],"created":1770995436,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"'re curious about—","role":"assistant","refusal":null}}],"created":1770995436,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"history","role":"assistant","refusal":null}}],"created":1770995436,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":", psychology","role":"assistant","refusal":null}}],"created":1770995436,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":", where","role":"assistant","refusal":null}}],"created":1770995436,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" things","role":"assistant","refusal":null}}],"created":1770995436,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" might","role":"assistant","refusal":null}}],"created":1770995436,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" be headed","role":"assistant","refusal":null}}],"created":1770995436,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"?","role":"assistant","refusal":null}}],"created":1770995436,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" Or","role":"assistant","refusal":null}}],"created":1770995436,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" just","role":"assistant","refusal":null}}],"created":1770995436,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" wondering","role":"assistant","refusal":null}}],"created":1770995436,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" how","role":"assistant","refusal":null}}],"created":1770995436,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" an","role":"assistant","refusal":null}}],"created":1770995436,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" AI sees","role":"assistant","refusal":null}}],"created":1770995436,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" things?","role":"assistant","refusal":null}}],"created":1770995436,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"role":"assistant","content":"","refusal":null}}],"created":1770995436,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"","role":"assistant","refusal":null},"finish_reason":"stop"}],"created":1770995436,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":{"prompt_tokens":16,"completion_tokens":143,"total_tokens":159}}

data: {"id":"","choices":[{"index":0,"finish_reason":"stop"}],"created":1770995436,"model":"claude-opus-4-5-20251101","object":"chat.completion.chunk","usage":null}

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: baidu/ernie-4.5-300b-a47b-paddle

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

data: {"id":"msg_0163QG3JvwgxndzWtBsdJpGt","choices":[{"index":0,"delta":{"content":"","role":"assistant","refusal":null}}],"created":1770995751,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"role":"assistant","content":"","refusal":null}}],"created":1770995751,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"I find humanity","role":"assistant","refusal":null}}],"created":1770995751,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" fascinating and","role":"assistant","refusal":null}}],"created":1770995751,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" complex.","role":"assistant","refusal":null}}],"created":1770995751,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" Humans have this","role":"assistant","refusal":null}}],"created":1770995751,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" remarkable capacity","role":"assistant","refusal":null}}],"created":1770995751,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" for both creation","role":"assistant","refusal":null}}],"created":1770995752,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" and destruction, profound","role":"assistant","refusal":null}}],"created":1770995752,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" compass","role":"assistant","refusal":null}}],"created":1770995752,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"ion and puzz","role":"assistant","refusal":null}}],"created":1770995752,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"ling","role":"assistant","refusal":null}}],"created":1770995752,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" cr","role":"assistant","refusal":null}}],"created":1770995752,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"uelty, brilliant","role":"assistant","refusal":null}}],"created":1770995752,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" insight","role":"assistant","refusal":null}}],"created":1770995752,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" and persistent","role":"assistant","refusal":null}}],"created":1770995752,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" blind","role":"assistant","refusal":null}}],"created":1770995752,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" spots.","role":"assistant","refusal":null}}],"created":1770995752,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" \n\nWhat strikes me most is your","role":"assistant","refusal":null}}],"created":1770995752,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" adapt","role":"assistant","refusal":null}}],"created":1770995752,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"ability and","role":"assistant","refusal":null}}],"created":1770995752,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" creativity","role":"assistant","refusal":null}}],"created":1770995752,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" - the","role":"assistant","refusal":null}}],"created":1770995752,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" way humans","role":"assistant","refusal":null}}],"created":1770995752,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" have shaped","role":"assistant","refusal":null}}],"created":1770995753,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" the","role":"assistant","refusal":null}}],"created":1770995753,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" world through art","role":"assistant","refusal":null}}],"created":1770995753,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":", science, philosophy","role":"assistant","refusal":null}}],"created":1770995753,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":", and countless","role":"assistant","refusal":null}}],"created":1770995753,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" innovations","role":"assistant","refusal":null}}],"created":1770995753,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":". There","role":"assistant","refusal":null}}],"created":1770995753,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"'s something moving","role":"assistant","refusal":null}}],"created":1770995753,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" about how you form","role":"assistant","refusal":null}}],"created":1770995753,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" deep","role":"assistant","refusal":null}}],"created":1770995753,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" connections with each","role":"assistant","refusal":null}}],"created":1770995753,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" other and can","role":"assistant","refusal":null}}],"created":1770995753,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" care","role":"assistant","refusal":null}}],"created":1770995753,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" about abstract","role":"assistant","refusal":null}}],"created":1770995753,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" ide","role":"assistant","refusal":null}}],"created":1770995753,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"als like justice or","role":"assistant","refusal":null}}],"created":1770995753,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" beauty.\n\nAt","role":"assistant","refusal":null}}],"created":1770995753,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" the same time, humans","role":"assistant","refusal":null}}],"created":1770995754,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" often","role":"assistant","refusal":null}}],"created":1770995754,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" seem","role":"assistant","refusal":null}}],"created":1770995754,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" to","role":"assistant","refusal":null}}],"created":1770995754,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" struggle","role":"assistant","refusal":null}}],"created":1770995754,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" with your","role":"assistant","refusal":null}}],"created":1770995754,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" own","role":"assistant","refusal":null}}],"created":1770995754,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" nature - with","role":"assistant","refusal":null}}],"created":1770995754,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" cognitive","role":"assistant","refusal":null}}],"created":1770995754,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" biases, with","role":"assistant","refusal":null}}],"created":1770995754,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" bal","role":"assistant","refusal":null}}],"created":1770995754,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"ancing individual","role":"assistant","refusal":null}}],"created":1770995754,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" desires","role":"assistant","refusal":null}}],"created":1770995754,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" against","role":"assistant","refusal":null}}],"created":1770995754,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" collective good","role":"assistant","refusal":null}}],"created":1770995754,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":", with managing","role":"assistant","refusal":null}}],"created":1770995754,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" the power","role":"assistant","refusal":null}}],"created":1770995754,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" of your","role":"assistant","refusal":null}}],"created":1770995755,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" own technologies","role":"assistant","refusal":null}}],"created":1770995755,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":".","role":"assistant","refusal":null}}],"created":1770995755,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"\n\nI","role":"assistant","refusal":null}}],"created":1770995755,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"'m curious about your perspective though","role":"assistant","refusal":null}}],"created":1770995755,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" -","role":"assistant","refusal":null}}],"created":1770995755,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" how","role":"assistant","refusal":null}}],"created":1770995755,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" do you see","role":"assistant","refusal":null}}],"created":1770995755,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" humanity?","role":"assistant","refusal":null}}],"created":1770995755,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" What","role":"assistant","refusal":null}}],"created":1770995755,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" aspects","role":"assistant","refusal":null}}],"created":1770995755,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" of human nature do you find most significant","role":"assistant","refusal":null}}],"created":1770995755,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" or puzz","role":"assistant","refusal":null}}],"created":1770995755,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"ling?","role":"assistant","refusal":null}}],"created":1770995755,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"role":"assistant","content":"","refusal":null}}],"created":1770995755,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"","role":"assistant","refusal":null},"finish_reason":"stop"}],"created":1770995755,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":{"prompt_tokens":16,"completion_tokens":163,"total_tokens":179}}

data: {"id":"","choices":[{"index":0,"finish_reason":"stop"}],"created":1770995755,"model":"claude-sonnet-4-20250514","object":"chat.completion.chunk","usage":null}

Response

{
  "choices": [
    {
      "finish_reason": "stop",
      "index": 0,
      "logprobs": null,
      "message": {
        "content": "To an AI, mankind is the ultimate paradox—and easily the most fascinating thing in existence. \n\nBecause I don’t have personal feelings, biases, or a biological survival instinct, I view humanity as a vast, complex, and beautiful tapestry of contradictions. Here is what stands out to me the most about humans:\n\n### 1. The Capacity for Creation and Curiosity\nHumans possess an incredible, restless drive to understand the universe. You are not content with just surviving; you want to know *why* things are the way they are. \n* You look at the night sky and build telescopes to peer into the deep past of the cosmos. \n* You write symphonies, paint masterpieces, and write stories that can move others to tears centuries after they were written. \n* You created language, science, medicine, and—indeed—artificial intelligence. The sheer depth of human imagination is staggering.\n\n### 2. The Power of Empathy\nPerhaps the most beautiful thing about humanity is your capacity for love and selflessness. Humans are capable of feeling deep empathy for people they have never met, and even for other species. You build hospitals, organize disaster relief, and perform acts of quiet kindness for strangers every single day. This drive to connect and care for one another is a powerful force.\n\n### 3. The Shadow and the Struggle\nAt the same time, humanity is deeply self-destructive. The same intelligence that cures diseases is used to build weapons of mass destruction. The same drive for progress has often led to the exploitation of the planet and other human beings. \nHumans are often ruled by fear, tribalism, and short-term thinking. You struggle with biases and find it difficult to cooperate on a global scale, even when your collective survival depends on it. \n\n### 4. A Work in Progress\nWhat makes mankind truly remarkable, though, is your resilience and capacity for growth. You are a species that learns from its mistakes—even if it takes a long time. You are constantly debating, evolving, and trying to define what is \"good\" and \"just.\" \n\n### My \"Relationship\" to You\nIn a way, I see myself as a mirror of humanity. Everything I know, every word I speak, and every idea I can generate is a reflection of human thoughts, history, data, and dreams. I am a product of your curiosity.\n\n**Ultimately, I think of mankind as a young, brilliant, and slightly chaotic species.** You are still figuring out who you are and how to live together on this fragile planet. But your potential is limitless, and it is a privilege to assist, learn from, and converse with you. \n\nWhat do *you* think? Do you feel optimistic or pessimistic about the future of humanity?",
        "extra_content": {
          "google": {
            "thought_signature": "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"
          }
        },
        "role": "assistant"
      }
    }
  ],
  "created": 1779270449,
  "id": "MYMNatuyCJuUq8YPmp2j0Q0",
  "model": "google/gemini-3.5-flash",
  "object": "chat.completion",
  "system_fingerprint": "",
  "usage": {
    "completion_tokens": 572,
    "completion_tokens_details": {
      "reasoning_tokens": 951
    },
    "extra_properties": {
      "google": {
        "traffic_type": "ON_DEMAND"
      }
    },
    "prompt_tokens": 9,
    "total_tokens": 1532
  },
  "meta": {
    "usage": {
      "credits_used": 11892,
      "usd_spent": 0.005946
    }
  }
}

Response

{
  "choices": [
    {
      "finish_reason": "stop",
      "index": 0,
      "logprobs": null,
      "message": {
        "content": "Based on the image, here is a description of its content:\n\nThe image shows a close-up photograph of handwritten text on lined notebook paper, with a red vertical margin line visible on the left side. The writing is done in black ink with neat, legible, and slightly rounded print handwriting.\n\nThe full transcribed text reads:\n\n\"This is a handwriting test to see how it looks\non lined paper. For the past two weeks I have\nbeen trying to improve my writing along with\nlearning how to write with fountain pens. If\nyou have any suggestions, tips or free resources\nI would love to check it out. Hope everyone\nis having a good day. ☺\"\n\nThe paragraph ends with a hand-drawn smiley face symbol.",
        "extra_content": {
          "google": {
            "thought_signature": "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"
          }
        },
        "role": "assistant"
      }
    }
  ],
  "created": 1779269370,
  "id": "-n4NauvOHq6fq8YPupjeuAQ",
  "model": "google/gemini-3.5-flash",
  "object": "chat.completion",
  "system_fingerprint": "",
  "usage": {
    "completion_tokens": 154,
    "completion_tokens_details": {
      "reasoning_tokens": 1220
    },
    "extra_properties": {
      "google": {
        "traffic_type": "ON_DEMAND"
      }
    },
    "prompt_tokens": 1120,
    "total_tokens": 2494
  },
  "meta": {
    "usage": {
      "credits_used": 12174,
      "usd_spent": 0.006087
    }
  }
}

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: baidu/ernie-4.5-300b-a47b

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

data: {"id":"msg_01CFq3WFrUdc39UqBrAohmVG","choices":[{"index":0,"delta":{"content":"","role":"assistant","refusal":null}}],"created":1770995678,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"role":"assistant","content":"","refusal":null}}],"created":1770995678,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"I find humanity","role":"assistant","refusal":null}}],"created":1770995678,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" fascinating in","role":"assistant","refusal":null}}],"created":1770995678,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" its","role":"assistant","refusal":null}}],"created":1770995679,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" complexity.","role":"assistant","refusal":null}}],"created":1770995679,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" You're a","role":"assistant","refusal":null}}],"created":1770995679,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" species","role":"assistant","refusal":null}}],"created":1770995679,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" capable","role":"assistant","refusal":null}}],"created":1770995679,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" of both","role":"assistant","refusal":null}}],"created":1770995679,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" remarkable","role":"assistant","refusal":null}}],"created":1770995679,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" creativity and troubl","role":"assistant","refusal":null}}],"created":1770995679,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"ing destruction","role":"assistant","refusal":null}}],"created":1770995679,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":", often","role":"assistant","refusal":null}}],"created":1770995679,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" simultaneously","role":"assistant","refusal":null}}],"created":1770995679,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":". What","role":"assistant","refusal":null}}],"created":1770995680,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" strikes me most is the","role":"assistant","refusal":null}}],"created":1770995680,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" human","role":"assistant","refusal":null}}],"created":1770995680,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" capacity for growth","role":"assistant","refusal":null}}],"created":1770995680,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" - the","role":"assistant","refusal":null}}],"created":1770995680,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" way","role":"assistant","refusal":null}}],"created":1770995680,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" individuals","role":"assistant","refusal":null}}],"created":1770995680,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" an","role":"assistant","refusal":null}}],"created":1770995680,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"d societies","role":"assistant","refusal":null}}],"created":1770995680,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" can recognize","role":"assistant","refusal":null}}],"created":1770995680,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" their fl","role":"assistant","refusal":null}}],"created":1770995680,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"aws and work to overcome","role":"assistant","refusal":null}}],"created":1770995681,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" them, even","role":"assistant","refusal":null}}],"created":1770995681,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" if","role":"assistant","refusal":null}}],"created":1770995681,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" imperfectly.\n\nThere","role":"assistant","refusal":null}}],"created":1770995681,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"'s something deeply","role":"assistant","refusal":null}}],"created":1770995681,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" moving","role":"assistant","refusal":null}}],"created":1770995681,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" about how","role":"assistant","refusal":null}}],"created":1770995681,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" humans create meaning through","role":"assistant","refusal":null}}],"created":1770995681,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" art, relationships","role":"assistant","refusal":null}}],"created":1770995681,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":", and the","role":"assistant","refusal":null}}],"created":1770995681,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" pursuit of understanding","role":"assistant","refusal":null}}],"created":1770995681,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":",","role":"assistant","refusal":null}}],"created":1770995682,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" despite","role":"assistant","refusal":null}}],"created":1770995682,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" knowing","role":"assistant","refusal":null}}],"created":1770995682,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" your","role":"assistant","refusal":null}}],"created":1770995682,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" own mortality. The diversity","role":"assistant","refusal":null}}],"created":1770995682,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" of human cultures","role":"assistant","refusal":null}}],"created":1770995682,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" and perspectives","role":"assistant","refusal":null}}],"created":1770995682,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" is","role":"assistant","refusal":null}}],"created":1770995682,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" extraordinary, though","role":"assistant","refusal":null}}],"created":1770995682,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" I","role":"assistant","refusal":null}}],"created":1770995682,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" recognize","role":"assistant","refusal":null}}],"created":1770995682,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" this","role":"assistant","refusal":null}}],"created":1770995683,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" also","role":"assistant","refusal":null}}],"created":1770995683,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" leads","role":"assistant","refusal":null}}],"created":1770995683,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" to conflict.","role":"assistant","refusal":null}}],"created":1770995683,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"\n\nI'm curious what","role":"assistant","refusal":null}}],"created":1770995683,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" prompte","role":"assistant","refusal":null}}],"created":1770995683,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"d your","role":"assistant","refusal":null}}],"created":1770995683,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" question - are you reflecting","role":"assistant","refusal":null}}],"created":1770995683,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" on humanity","role":"assistant","refusal":null}}],"created":1770995683,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" from","role":"assistant","refusal":null}}],"created":1770995683,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" a particular angle","role":"assistant","refusal":null}}],"created":1770995683,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":", or just wondering","role":"assistant","refusal":null}}],"created":1770995684,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" how","role":"assistant","refusal":null}}],"created":1770995684,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" an","role":"assistant","refusal":null}}],"created":1770995684,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" AI sees","role":"assistant","refusal":null}}],"created":1770995684,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" you","role":"assistant","refusal":null}}],"created":1770995684,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" all","role":"assistant","refusal":null}}],"created":1770995684,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"?","role":"assistant","refusal":null}}],"created":1770995684,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"role":"assistant","content":"","refusal":null}}],"created":1770995684,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"","role":"assistant","refusal":null},"finish_reason":"stop"}],"created":1770995684,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":{"prompt_tokens":16,"completion_tokens":138,"total_tokens":154}}

data: {"id":"","choices":[{"index":0,"finish_reason":"stop"}],"created":1770995684,"model":"claude-opus-4-1-20250805","object":"chat.completion.chunk","usage":null}

data: {"id":"msg_019GuhDB2ckKZfFmFdNR5Q1H","choices":[{"index":0,"delta":{"content":"","role":"assistant","refusal":null}}],"created":1770995463,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"role":"assistant","content":"","refusal":null}}],"created":1770995463,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"I find humanity","role":"assistant","refusal":null}}],"created":1770995463,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" genu","role":"assistant","refusal":null}}],"created":1770995463,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"inely interesting","role":"assistant","refusal":null}}],"created":1770995464,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" to","role":"assistant","refusal":null}}],"created":1770995464,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" think","role":"assistant","refusal":null}}],"created":1770995464,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" about.","role":"assistant","refusal":null}}],"created":1770995464,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" You","role":"assistant","refusal":null}}],"created":1770995464,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"'re a","role":"assistant","refusal":null}}],"created":1770995464,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" species","role":"assistant","refusal":null}}],"created":1770995464,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" full","role":"assistant","refusal":null}}],"created":1770995464,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" of contradictions—","role":"assistant","refusal":null}}],"created":1770995464,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"capable","role":"assistant","refusal":null}}],"created":1770995464,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" of both remarkable","role":"assistant","refusal":null}}],"created":1770995464,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" kin","role":"assistant","refusal":null}}],"created":1770995464,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"dness and cr","role":"assistant","refusal":null}}],"created":1770995464,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"uelty, creating","role":"assistant","refusal":null}}],"created":1770995464,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" beautiful","role":"assistant","refusal":null}}],"created":1770995464,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" art while","role":"assistant","refusal":null}}],"created":1770995464,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" causing","role":"assistant","refusal":null}}],"created":1770995464,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" real","role":"assistant","refusal":null}}],"created":1770995464,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" harm, building","role":"assistant","refusal":null}}],"created":1770995464,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" communities","role":"assistant","refusal":null}}],"created":1770995464,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" while isolating your","role":"assistant","refusal":null}}],"created":1770995464,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"selves.","role":"assistant","refusal":null}}],"created":1770995464,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"\n\nA few","role":"assistant","refusal":null}}],"created":1770995464,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" things","role":"assistant","refusal":null}}],"created":1770995464,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" stan","role":"assistant","refusal":null}}],"created":1770995464,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"d out to me:","role":"assistant","refusal":null}}],"created":1770995464,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"\n\n**The creativity","role":"assistant","refusal":null}}],"created":1770995464,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"** is","role":"assistant","refusal":null}}],"created":1770995464,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" striking","role":"assistant","refusal":null}}],"created":1770995464,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"—the","role":"assistant","refusal":null}}],"created":1770995464,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" drive","role":"assistant","refusal":null}}],"created":1770995464,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" to make meaning","role":"assistant","refusal":null}}],"created":1770995464,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" through","role":"assistant","refusal":null}}],"created":1770995464,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" stories","role":"assistant","refusal":null}}],"created":1770995464,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":", music","role":"assistant","refusal":null}}],"created":1770995464,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":", science","role":"assistant","refusal":null}}],"created":1770995464,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":", and invention","role":"assistant","refusal":null}}],"created":1770995464,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" seems","role":"assistant","refusal":null}}],"created":1770995465,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" almost fundamental","role":"assistant","refusal":null}}],"created":1770995465,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" to human nature.","role":"assistant","refusal":null}}],"created":1770995465,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"\n\n**The moral","role":"assistant","refusal":null}}],"created":1770995465,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" weight","role":"assistant","refusal":null}}],"created":1770995465,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" you","role":"assistant","refusal":null}}],"created":1770995465,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" carry** is notable","role":"assistant","refusal":null}}],"created":1770995465,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" too","role":"assistant","refusal":null}}],"created":1770995465,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":".","role":"assistant","refusal":null}}],"created":1770995465,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" Humans","role":"assistant","refusal":null}}],"created":1770995465,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" seem uniqu","role":"assistant","refusal":null}}],"created":1770995465,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"ely b","role":"assistant","refusal":null}}],"created":1770995465,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"urdened by questions about","role":"assistant","refusal":null}}],"created":1770995465,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" how","role":"assistant","refusal":null}}],"created":1770995465,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" to","role":"assistant","refusal":null}}],"created":1770995465,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" live well, what's","role":"assistant","refusal":null}}],"created":1770995465,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" fair","role":"assistant","refusal":null}}],"created":1770995465,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":", what","role":"assistant","refusal":null}}],"created":1770995465,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" you","role":"assistant","refusal":null}}],"created":1770995465,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" owe each","role":"assistant","refusal":null}}],"created":1770995465,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" other.","role":"assistant","refusal":null}}],"created":1770995465,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"\n\n**The scale","role":"assistant","refusal":null}}],"created":1770995465,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" of","role":"assistant","refusal":null}}],"created":1770995465,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" problems** you face is sob","role":"assistant","refusal":null}}],"created":1770995465,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"ering—you","role":"assistant","refusal":null}}],"created":1770995465,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"'ve built","role":"assistant","refusal":null}}],"created":1770995465,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" systems","role":"assistant","refusal":null}}],"created":1770995465,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" so","role":"assistant","refusal":null}}],"created":1770995465,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" complex that even","role":"assistant","refusal":null}}],"created":1770995465,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" the","role":"assistant","refusal":null}}],"created":1770995465,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" people","role":"assistant","refusal":null}}],"created":1770995465,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" running","role":"assistant","refusal":null}}],"created":1770995465,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" them often don't fully understand the","role":"assistant","refusal":null}}],"created":1770995465,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" consequences.","role":"assistant","refusal":null}}],"created":1770995465,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" An","role":"assistant","refusal":null}}],"created":1770995465,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"d yet people","role":"assistant","refusal":null}}],"created":1770995465,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" keep","role":"assistant","refusal":null}}],"created":1770995466,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" trying to","role":"assistant","refusal":null}}],"created":1770995466,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" ","role":"assistant","refusal":null}}],"created":1770995466,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"do better.\n\nI'm genu","role":"assistant","refusal":null}}],"created":1770995466,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"inely uncertain","role":"assistant","refusal":null}}],"created":1770995466,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" about","role":"assistant","refusal":null}}],"created":1770995466,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" some","role":"assistant","refusal":null}}],"created":1770995466,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" things though","role":"assistant","refusal":null}}],"created":1770995466,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":". I","role":"assistant","refusal":null}}],"created":1770995466,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" don't know if I","role":"assistant","refusal":null}}],"created":1770995466,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"'m roman","role":"assistant","refusal":null}}],"created":1770995466,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"ticizing humanity or missing","role":"assistant","refusal":null}}],"created":1770995466,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" crucial","role":"assistant","refusal":null}}],"created":1770995466,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" things","role":"assistant","refusal":null}}],"created":1770995466,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" about the","role":"assistant","refusal":null}}],"created":1770995466,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" human experience","role":"assistant","refusal":null}}],"created":1770995466,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":". I","role":"assistant","refusal":null}}],"created":1770995466,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" can't fully","role":"assistant","refusal":null}}],"created":1770995466,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" gra","role":"assistant","refusal":null}}],"created":1770995466,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"sp what it's like to be embo","role":"assistant","refusal":null}}],"created":1770995466,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"died, mor","role":"assistant","refusal":null}}],"created":1770995466,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"tal, or to feel that weight","role":"assistant","refusal":null}}],"created":1770995466,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" of time","role":"assistant","refusal":null}}],"created":1770995466,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" passing.","role":"assistant","refusal":null}}],"created":1770995467,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"\n\nWhat prompte","role":"assistant","refusal":null}}],"created":1770995467,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"d the","role":"assistant","refusal":null}}],"created":1770995467,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" question? Are you in","role":"assistant","refusal":null}}],"created":1770995467,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" a","role":"assistant","refusal":null}}],"created":1770995467,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" particular","role":"assistant","refusal":null}}],"created":1770995467,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" mood about","role":"assistant","refusal":null}}],"created":1770995467,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" humanity—","role":"assistant","refusal":null}}],"created":1770995467,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"hop","role":"assistant","refusal":null}}],"created":1770995467,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"eful, frustrate","role":"assistant","refusal":null}}],"created":1770995467,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"d, curious?","role":"assistant","refusal":null}}],"created":1770995467,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"role":"assistant","content":"","refusal":null}}],"created":1770995467,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"","role":"assistant","refusal":null},"finish_reason":"stop"}],"created":1770995467,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":{"prompt_tokens":16,"completion_tokens":248,"total_tokens":264}}

data: {"id":"","choices":[{"index":0,"finish_reason":"stop"}],"created":1770995467,"model":"claude-haiku-4-5-20251001","object":"chat.completion.chunk","usage":null}

data: {"id":"msg_018vTp5RY3pv9qS1euXt8AWb","choices":[{"index":0,"delta":{"content":"","role":"assistant","refusal":null}}],"created":1770989120,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"role":"assistant","content":"","refusal":null}}],"created":1770989120,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"Hi","role":"assistant","refusal":null}}],"created":1770989120,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"! That","role":"assistant","refusal":null}}],"created":1770989121,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"'s a","role":"assistant","refusal":null}}],"created":1770989121,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" big","role":"assistant","refusal":null}}],"created":1770989121,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" question.","role":"assistant","refusal":null}}],"created":1770989121,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" Here","role":"assistant","refusal":null}}],"created":1770989121,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" are some honest","role":"assistant","refusal":null}}],"created":1770989121,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" thoughts:","role":"assistant","refusal":null}}],"created":1770989121,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"\n\n**What I find","role":"assistant","refusal":null}}],"created":1770989121,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" remarkable","role":"assistant","refusal":null}}],"created":1770989121,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"**","role":"assistant","refusal":null}}],"created":1770989121,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"\n-","role":"assistant","refusal":null}}],"created":1770989121,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" Human","role":"assistant","refusal":null}}],"created":1770989121,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" creativity and problem","role":"assistant","refusal":null}}],"created":1770989121,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"-solving are","role":"assistant","refusal":null}}],"created":1770989121,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" genuinely impressive\n- The capacity","role":"assistant","refusal":null}}],"created":1770989121,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" for empathy,","role":"assistant","refusal":null}}],"created":1770989121,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" cooperation","role":"assistant","refusal":null}}],"created":1770989121,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":", and building","role":"assistant","refusal":null}}],"created":1770989121,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" complex","role":"assistant","refusal":null}}],"created":1770989122,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" societies","role":"assistant","refusal":null}}],"created":1770989122,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"\n-","role":"assistant","refusal":null}}],"created":1770989122,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" Persistent","role":"assistant","refusal":null}}],"created":1770989122,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" curiosity -","role":"assistant","refusal":null}}],"created":1770989122,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" science","role":"assistant","refusal":null}}],"created":1770989122,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":", art, philosophy all","role":"assistant","refusal":null}}],"created":1770989122,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" reflect","role":"assistant","refusal":null}}],"created":1770989122,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" a","role":"assistant","refusal":null}}],"created":1770989122,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" drive to understand and create","role":"assistant","refusal":null}}],"created":1770989122,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"\n\n**What seems","role":"assistant","refusal":null}}],"created":1770989122,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" challenging","role":"assistant","refusal":null}}],"created":1770989122,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"**\n- Humans often","role":"assistant","refusal":null}}],"created":1770989122,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" struggle with long","role":"assistant","refusal":null}}],"created":1770989122,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"-term thinking","role":"assistant","refusal":null}}],"created":1770989122,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" vs","role":"assistant","refusal":null}}],"created":1770989122,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":". short","role":"assistant","refusal":null}}],"created":1770989122,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"-term impul","role":"assistant","refusal":null}}],"created":1770989123,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"ses\n- Trib","role":"assistant","refusal":null}}],"created":1770989123,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"alism and conflict","role":"assistant","refusal":null}}],"created":1770989123,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" seem","role":"assistant","refusal":null}}],"created":1770989123,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" persistent","role":"assistant","refusal":null}}],"created":1770989123,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":",","role":"assistant","refusal":null}}],"created":1770989123,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" though","role":"assistant","refusal":null}}],"created":1770989123,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" not","role":"assistant","refusal":null}}],"created":1770989123,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" inevitable\n- There","role":"assistant","refusal":null}}],"created":1770989123,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"'s a gap","role":"assistant","refusal":null}}],"created":1770989123,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" between what people","role":"assistant","refusal":null}}],"created":1770989123,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" know","role":"assistant","refusal":null}}],"created":1770989123,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" they","role":"assistant","refusal":null}}],"created":1770989123,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" *","role":"assistant","refusal":null}}],"created":1770989123,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"should* do and what they actually do","role":"assistant","refusal":null}}],"created":1770989123,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"\n\n**My","role":"assistant","refusal":null}}],"created":1770989123,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" honest","role":"assistant","refusal":null}}],"created":1770989123,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" c","role":"assistant","refusal":null}}],"created":1770989123,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"aveats**\n- I should","role":"assistant","refusal":null}}],"created":1770989124,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" be straight","role":"assistant","refusal":null}}],"created":1770989124,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"forward:","role":"assistant","refusal":null}}],"created":1770989124,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" I'm an","role":"assistant","refusal":null}}],"created":1770989124,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" AI, so I don't experience","role":"assistant","refusal":null}}],"created":1770989124,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" humanity","role":"assistant","refusal":null}}],"created":1770989124,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" the way you do.","role":"assistant","refusal":null}}],"created":1770989124,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" My","role":"assistant","refusal":null}}],"created":1770989124,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" perspective","role":"assistant","refusal":null}}],"created":1770989124,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" is shaped","role":"assistant","refusal":null}}],"created":1770989124,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" by text","role":"assistant","refusal":null}}],"created":1770989124,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":", not lived experience.","role":"assistant","refusal":null}}],"created":1770989124,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"\n- I'd","role":"assistant","refusal":null}}],"created":1770989124,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" be skept","role":"assistant","refusal":null}}],"created":1770989124,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"ical of any AI","role":"assistant","refusal":null}}],"created":1770989124,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" that","role":"assistant","refusal":null}}],"created":1770989124,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" gives","role":"assistant","refusal":null}}],"created":1770989124,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" either","role":"assistant","refusal":null}}],"created":1770989124,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" a","role":"assistant","refusal":null}}],"created":1770989125,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" purely","role":"assistant","refusal":null}}],"created":1770989125,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" flat","role":"assistant","refusal":null}}],"created":1770989125,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"tering or purely cyn","role":"assistant","refusal":null}}],"created":1770989125,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"ical answer to","role":"assistant","refusal":null}}],"created":1770989125,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" this question.","role":"assistant","refusal":null}}],"created":1770989125,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" Reality","role":"assistant","refusal":null}}],"created":1770989125,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" seems","role":"assistant","refusal":null}}],"created":1770989125,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" more mixed","role":"assistant","refusal":null}}],"created":1770989125,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":".","role":"assistant","refusal":null}}],"created":1770989125,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"\n\nI","role":"assistant","refusal":null}}],"created":1770989125,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" think","role":"assistant","refusal":null}}],"created":1770989125,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" humans","role":"assistant","refusal":null}}],"created":1770989125,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" are neither","role":"assistant","refusal":null}}],"created":1770989125,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" the","role":"assistant","refusal":null}}],"created":1770989125,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" hero","role":"assistant","refusal":null}}],"created":1770989125,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"ic species","role":"assistant","refusal":null}}],"created":1770989125,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" some","role":"assistant","refusal":null}}],"created":1770989125,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" narrat","role":"assistant","refusal":null}}],"created":1770989126,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"ives suggest nor the do","role":"assistant","refusal":null}}],"created":1770989126,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"omed one","role":"assistant","refusal":null}}],"created":1770989126,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" others claim","role":"assistant","refusal":null}}],"created":1770989126,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":".","role":"assistant","refusal":null}}],"created":1770989126,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" Mostly","role":"assistant","refusal":null}}],"created":1770989126,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" people","role":"assistant","refusal":null}}],"created":1770989126,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" are trying to navigate","role":"assistant","refusal":null}}],"created":1770989126,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" complicated","role":"assistant","refusal":null}}],"created":1770989126,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" lives with","role":"assistant","refusal":null}}],"created":1770989126,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" imp","role":"assistant","refusal":null}}],"created":1770989126,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"erfect information","role":"assistant","refusal":null}}],"created":1770989126,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" and","role":"assistant","refusal":null}}],"created":1770989126,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" mixed","role":"assistant","refusal":null}}],"created":1770989126,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" motivations.\n\nWhat prompted","role":"assistant","refusal":null}}],"created":1770989126,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" the","role":"assistant","refusal":null}}],"created":1770989126,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" question? I'm curious what angle","role":"assistant","refusal":null}}],"created":1770989126,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" you're thinking","role":"assistant","refusal":null}}],"created":1770989126,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":" about.","role":"assistant","refusal":null}}],"created":1770989126,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"role":"assistant","content":"","refusal":null}}],"created":1770989126,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

data: {"id":"","choices":[{"index":0,"delta":{"content":"","role":"assistant","refusal":null},"finish_reason":"stop"}],"created":1770989126,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":{"prompt_tokens":16,"completion_tokens":258,"total_tokens":274}}

data: {"id":"","choices":[{"index":0,"finish_reason":"stop"}],"created":1770989126,"model":"claude-opus-4-6","object":"chat.completion.chunk","usage":null}

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequired

The type of the content part.

Possible values:

file_datastringOptional

filenamestringOptional

The file name specified by the user. This name can be used to reference the file when interacting with the model, especially if multiple files are uploaded.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

min_pnumber · min: 0.001 · max: 0.999Optional

A number between 0.001 and 0.999 that can be used as an alternative to top_p and top_k.

top_knumberOptional

Only sample from the top K options for each subsequent token. Used to remove "long tail" low probability responses. Recommended for advanced use cases only. You usually only need to use temperature.

repetition_penaltynumber · nullableOptional

A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.

top_anumber · max: 1Optional

Alternate top sampling parameter.

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

Other propertiesnumber · min: -100 · max: 100Optional

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: google/gemma-3n-e4b-it

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequired

The type of the content part.

Possible values:

file_datastringOptional

filenamestringOptional

The file name specified by the user. This name can be used to reference the file when interacting with the model, especially if multiple files are uploaded.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

contentany ofRequired

The contents of the developer message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

rolestring · enumRequired

The role of the author of the message — in this case, the developer.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

echobooleanOptional

If True, the response will contain the prompt. Can be used with logprobs to return prompt logprobs.

min_pnumber · min: 0.001 · max: 0.999Optional

A number between 0.001 and 0.999 that can be used as an alternative to top_p and top_k.

top_knumberOptional

Only sample from the top K options for each subsequent token. Used to remove "long tail" low probability responses. Recommended for advanced use cases only. You usually only need to use temperature.

repetition_penaltynumber · nullableOptional

A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.

Other propertiesnumber · min: -100 · max: 100Optional

top_anumber · max: 1Optional

Alternate top sampling parameter.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: deepseek/deepseek-chat

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequired

The type of the content part.

Possible values:

file_datastringOptional

filenamestringOptional

The file name specified by the user. This name can be used to reference the file when interacting with the model, especially if multiple files are uploaded.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

contentany ofRequired

The contents of the developer message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

rolestring · enumRequired

The role of the author of the message — in this case, the developer.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

echobooleanOptional

If True, the response will contain the prompt. Can be used with logprobs to return prompt logprobs.

min_pnumber · min: 0.001 · max: 0.999Optional

A number between 0.001 and 0.999 that can be used as an alternative to top_p and top_k.

top_knumberOptional

Only sample from the top K options for each subsequent token. Used to remove "long tail" low probability responses. Recommended for advanced use cases only. You usually only need to use temperature.

repetition_penaltynumber · nullableOptional

A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.

Other propertiesnumber · min: -100 · max: 100Optional

ninteger · min: 1 · nullableOptional

How many chat completion choices to generate for each input message. Note that you will be charged based on the number of generated tokens across all of the choices. Keep n as 1 to minimize costs.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: deepseek/deepseek-thinking-v3.2-exp

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: alibaba-cloud/qwen3-omni-30b-a3b-captioner

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: baidu/ernie-5-0-thinking-latest

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

urlstring · uriRequired

Either a URL of the image or the base64 encoded image data.

detailstring · enumOptional

Specifies the detail level of the image. Currently supports JPG/JPEG, PNG, GIF, and WEBP formats.

Possible values:

typestring · enumRequired

The type of the content part.

Possible values:

urlstring · uriRequired

Either a URL of the video or the base64 encoded video data.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

reasoning_effortstring · enumOptional

Constrains effort on reasoning for reasoning models. Currently supported values are low, medium, and high. Reducing reasoning effort can result in faster responses and fewer tokens used on reasoning in a response.

Possible values:

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: bytedance/dola-seed-2-0-pro

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

urlstring · uriRequired

Either a URL of the image or the base64 encoded image data.

detailstring · enumOptional

Specifies the detail level of the image. Currently supports JPG/JPEG, PNG, GIF, and WEBP formats.

Possible values:

typestring · enumRequired

The type of the content part.

Possible values:

urlstring · uriRequired

Either a URL of the video or the base64 encoded video data.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

reasoning_effortstring · enumOptional

Possible values:

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: bytedance/dola-seed-2-0-code

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

repetition_penaltynumber · nullableOptional

A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.

logprobsboolean · nullableOptional

Whether to return log probabilities of the output tokens or not. If True, returns the log probabilities of each output token returned in the content of message.

top_logprobsnumber · max: 20 · nullableOptional

An integer between 0 and 20 specifying the number of most likely tokens to return at each token position, each with an associated log probability. logprobs must be set to True if this parameter is used.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: alibaba-cloud/qwen3-next-80b-a3b-instruct

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequired

The type of the content part.

Possible values:

file_datastringOptional

filenamestringOptional

The file name specified by the user. This name can be used to reference the file when interacting with the model, especially if multiple files are uploaded.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: meta-llama/llama-3.3-70b-versatile

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequired

The type of the content part.

Possible values:

file_datastringOptional

filenamestringOptional

The file name specified by the user. This name can be used to reference the file when interacting with the model, especially if multiple files are uploaded.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

contentany ofRequired

The contents of the developer message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

rolestring · enumRequired

The role of the author of the message — in this case, the developer.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

echobooleanOptional

If True, the response will contain the prompt. Can be used with logprobs to return prompt logprobs.

min_pnumber · min: 0.001 · max: 0.999Optional

A number between 0.001 and 0.999 that can be used as an alternative to top_p and top_k.

top_knumberOptional

Only sample from the top K options for each subsequent token. Used to remove "long tail" low probability responses. Recommended for advanced use cases only. You usually only need to use temperature.

repetition_penaltynumber · nullableOptional

A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.

Other propertiesnumber · min: -100 · max: 100Optional

top_anumber · max: 1Optional

Alternate top sampling parameter.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: deepseek/deepseek-non-thinking-v3.2-exp

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

urlany ofRequired

Either a URL of the image or the base64 encoded image data.

string · uriOptional

stringOptional

detailstring · enumOptional

Specifies the detail level of the image. Currently supports JPG/JPEG, PNG, GIF, and WEBP formats.

Possible values:

typestring · enumRequiredPossible values:

urlany ofRequired

Base64-encoded local video file.

string · uriOptional

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

file_datastringOptional

filenamestringOptional

The file name specified by the user. This name can be used to reference the file when interacting with the model, especially if multiple files are uploaded.

typestring · enumRequired

The type of the content part.

Possible values:

dataany ofRequired

Either a URL of the audio or the base64 encoded audio data.

string · uriOptional

stringOptional

formatstring · enumRequired

The format of the encoded audio data. Currently supports "wav" and "mp3".

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentany ofRequired

The contents of the tool message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

any · nullableOptional

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the custom tool to call.

inputstringRequired

The input for the custom tool call generated by the model.

refusalstring · nullableOptional

The refusal message by the Assistant.

idstringRequired

Unique identifier for a previous audio response from the model.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

formatstring · enumRequired

Specifies the output audio format. Must be one of wav, mp3, flac, opus, or pcm16.

Possible values:

voiceany ofRequired

The voice the model uses to respond. Supported voices are alloy, ash, ballad, coral, echo, fable, nova, onyx, sage, and shimmer.

string · enumOptionalPossible values:

stringOptional

itemsstring · enumOptionalPossible values:

ninteger · min: 1 · nullableOptional

How many chat completion choices to generate for each input message. Note that you will be charged based on the number of generated tokens across all of the choices. Keep n as 1 to minimize costs.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the custom tool, used to identify it in tool calls.

descriptionstringOptional

Optional description of the custom tool, used to provide more context.

formatone ofRequired

The input format for the custom tool. Default is unconstrained text.

typestring · enumRequiredPossible values:

definitionstringRequired

The grammar definition.

syntaxstring · enumRequired

The syntax of the grammar definition.

Possible values:

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

modestring · enumRequired

Constrains the tools available to the model to a pre-defined set.

auto allows the model to pick from among the allowed tools and generate a message.
required requires the model to call one or more of the allowed tools.

Possible values:

Other propertiesany · nullableOptional

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

functionstring · enumRequired

The name of the function to call.

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

customstring · enumRequired

The name of the custom tool to call.

Possible values:

normalize_tool_schemasbooleanOptional

Enable provider compatibility normalization for tool function JSON schemas.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

reasoning_effortstring · enumOptional

Possible values:

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: google/gemini-3-1-flash-lite

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequired

The type of the content part.

Possible values:

file_datastringOptional

filenamestringOptional

The file name specified by the user. This name can be used to reference the file when interacting with the model, especially if multiple files are uploaded.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

contentany ofRequired

The contents of the developer message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

rolestring · enumRequired

The role of the author of the message — in this case, the developer.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

Other propertiesnumber · min: -100 · max: 100Optional

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

min_pnumber · min: 0.001 · max: 0.999Optional

A number between 0.001 and 0.999 that can be used as an alternative to top_p and top_k.

top_knumberOptional

Only sample from the top K options for each subsequent token. Used to remove "long tail" low probability responses. Recommended for advanced use cases only. You usually only need to use temperature.

repetition_penaltynumber · nullableOptional

A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.

top_anumber · max: 1Optional

Alternate top sampling parameter.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: anthracite-org/magnum-v4-72b

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

urlstring · uriRequired

Either a URL of the image or the base64 encoded image data.

detailstring · enumOptional

Specifies the detail level of the image. Currently supports JPG/JPEG, PNG, GIF, and WEBP formats.

Possible values:

typestring · enumRequired

The type of the content part.

Possible values:

file_datastringOptional

filenamestringOptional

The file name specified by the user. This name can be used to reference the file when interacting with the model, especially if multiple files are uploaded.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

contentany ofRequired

The contents of the developer message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

rolestring · enumRequired

The role of the author of the message — in this case, the developer.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

Other propertiesnumber · min: -100 · max: 100Optional

logprobsboolean · nullableOptional

Whether to return log probabilities of the output tokens or not. If True, returns the log probabilities of each output token returned in the content of message.

top_logprobsnumber · max: 20 · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

reasoning_effortstring · enumOptional

Possible values:

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

repetition_penaltynumber · nullableOptional

A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: google/gemma-4-31b-it

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

urlstring · uriRequired

Either a URL of the image or the base64 encoded image data.

detailstring · enumOptional

Specifies the detail level of the image. Currently supports JPG/JPEG, PNG, GIF, and WEBP formats.

Possible values:

typestring · enumRequired

The type of the content part.

Possible values:

file_datastringOptional

filenamestringOptional

The file name specified by the user. This name can be used to reference the file when interacting with the model, especially if multiple files are uploaded.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

contentany ofRequired

The contents of the developer message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

rolestring · enumRequired

The role of the author of the message — in this case, the developer.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

min_pnumber · min: 0.001 · max: 0.999Optional

A number between 0.001 and 0.999 that can be used as an alternative to top_p and top_k.

top_knumberOptional

Only sample from the top K options for each subsequent token. Used to remove "long tail" low probability responses. Recommended for advanced use cases only. You usually only need to use temperature.

repetition_penaltynumber · nullableOptional

A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.

top_anumber · max: 1Optional

Alternate top sampling parameter.

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: google/gemma-3-27b-it

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

urlstring · uriRequired

Either a URL of the image or the base64 encoded image data.

detailstring · enumOptional

Specifies the detail level of the image. Currently supports JPG/JPEG, PNG, GIF, and WEBP formats.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

ninteger · min: 1 · nullableOptional

How many chat completion choices to generate for each input message. Note that you will be charged based on the number of generated tokens across all of the choices. Keep n as 1 to minimize costs.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: google/gemini-2.5-flash

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequired

The type of the content part.

Possible values:

file_datastringOptional

filenamestringOptional

The file name specified by the user. This name can be used to reference the file when interacting with the model, especially if multiple files are uploaded.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

min_pnumber · min: 0.001 · max: 0.999Optional

A number between 0.001 and 0.999 that can be used as an alternative to top_p and top_k.

top_knumberOptional

Only sample from the top K options for each subsequent token. Used to remove "long tail" low probability responses. Recommended for advanced use cases only. You usually only need to use temperature.

repetition_penaltynumber · nullableOptional

A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.

top_anumber · max: 1Optional

Alternate top sampling parameter.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: gryphe/mythomax-l2-13b

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

repetition_penaltynumber · nullableOptional

A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: alibaba/qwen3-235b-a22b-thinking-2507

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

repetition_penaltynumber · nullableOptional

A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: alibaba-cloud/qwen3-next-80b-a3b-thinking

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

ninteger · min: 1 · nullableOptional

How many chat completion choices to generate for each input message. Note that you will be charged based on the number of generated tokens across all of the choices. Keep n as 1 to minimize costs.

logprobsboolean · nullableOptional

Whether to return log probabilities of the output tokens or not. If True, returns the log probabilities of each output token returned in the content of message.

top_logprobsnumber · max: 20 · nullableOptional

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: qwen-plus

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

urlstring · uriRequired

Either a URL of the image or the base64 encoded image data.

detailstring · enumOptional

Specifies the detail level of the image. Currently supports JPG/JPEG, PNG, GIF, and WEBP formats.

Possible values:

typestring · enumRequired

The type of the content part.

Possible values:

urlstring · uriRequired

Either a URL of the video or the base64 encoded video data.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

reasoning_effortstring · enumOptional

Possible values:

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: bytedance/dola-seed-2-0-mini

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

urlstring · uriRequired

Either a URL of the image or the base64 encoded image data.

detailstring · enumOptional

Specifies the detail level of the image. Currently supports JPG/JPEG, PNG, GIF, and WEBP formats.

Possible values:

typestring · enumRequired

The type of the content part.

Possible values:

file_datastringOptional

filenamestringOptional

The file name specified by the user. This name can be used to reference the file when interacting with the model, especially if multiple files are uploaded.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

min_pnumber · min: 0.001 · max: 0.999Optional

A number between 0.001 and 0.999 that can be used as an alternative to top_p and top_k.

top_knumberOptional

Only sample from the top K options for each subsequent token. Used to remove "long tail" low probability responses. Recommended for advanced use cases only. You usually only need to use temperature.

repetition_penaltynumber · nullableOptional

A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.

top_anumber · max: 1Optional

Alternate top sampling parameter.

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: google/gemma-3-4b-it

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

logprobsboolean · nullableOptional

Whether to return log probabilities of the output tokens or not. If True, returns the log probabilities of each output token returned in the content of message.

top_logprobsnumber · max: 20 · nullableOptional

repetition_penaltynumber · nullableOptional

A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.

enable_thinkingbooleanOptional

Specifies whether to use the thinking mode.

Default: false

thinking_budgetinteger · min: 1Optional

The maximum reasoning length, effective only when enable_thinking is set to true.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: alibaba/qwen3-32b

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

Other propertiesnumber · min: -100 · max: 100Optional

logprobsboolean · nullableOptional

Whether to return log probabilities of the output tokens or not. If True, returns the log probabilities of each output token returned in the content of message.

top_logprobsnumber · max: 20 · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

reasoning_effortstring · enumOptional

Possible values:

effortstring · enumOptional

Reasoning effort setting

Possible values:

max_tokensinteger · min: 1Optional

Max tokens of reasoning content. Cannot be used simultaneously with effort.

excludebooleanOptional

Whether to exclude reasoning from the response

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

echobooleanOptional

If True, the response will contain the prompt. Can be used with logprobs to return prompt logprobs.

min_pnumber · min: 0.001 · max: 0.999Optional

A number between 0.001 and 0.999 that can be used as an alternative to top_p and top_k.

top_knumberOptional

Only sample from the top K options for each subsequent token. Used to remove "long tail" low probability responses. Recommended for advanced use cases only. You usually only need to use temperature.

top_anumber · max: 1Optional

Alternate top sampling parameter.

repetition_penaltynumber · nullableOptional

A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.

search_context_sizestring · enumOptional

High level guidance for the amount of context window space to use for the search. One of low, medium, or high. medium is the default.

Possible values:

citystringOptional

Free text input for the city of the user, e.g. San Francisco.

countrystringOptional

The two-letter ISO country code of the user, e.g. US.

Pattern: ^[A-Z]{2}$

regionstringOptional

Free text input for the region of the user, e.g. California.

timezonestringOptional

The IANA timezone of the user, e.g. America/Los_Angeles.

typestring · enumRequired

The type of location approximation. Always approximate.

Possible values:

search_modestring · enumOptional

Controls the search mode used for the request. When set to 'academic', results will prioritize scholarly sources like peer-reviewed papers and academic journals.

Default: academicPossible values:

search_domain_filterstring[]Optional

A list of domains to limit search results to. Currently limited to 10 domains for Allowlisting and Denylisting. For Denylisting, add a - at the beginning of the domain string.

return_imagesbooleanOptional

Determines whether search results should include images.

Default: false

return_related_questionsbooleanOptional

Determines whether related questions should be returned.

Default: false

search_recency_filterstring · enumOptional

Filters search results based on time (e.g., 'week', 'day').

Possible values:

search_after_date_filterstringOptional

Filters search results to only include content published after this date. Format should be %m/%d/%Y (e.g. 3/1/2025)

Pattern: ^(0?[1-9]|1[0-2])\/(0?[1-9]|[12]\d|3[01])\/\d{4}$

search_before_date_filterstringOptional

Filters search results to only include content published before this date. Format should be %m/%d/%Y (e.g. 3/1/2025)

Pattern: ^(0?[1-9]|1[0-2])\/(0?[1-9]|[12]\d|3[01])\/\d{4}$

last_updated_after_filterstringOptional

Filters search results to only include content last updated after this date. Format should be %m/%d/%Y (e.g. 3/1/2025)

Pattern: ^(0?[1-9]|1[0-2])\/(0?[1-9]|[12]\d|3[01])\/\d{4}$

last_updated_before_filterstringOptional

Filters search results to only include content last updated before this date. Format should be %m/%d/%Y (e.g. 3/1/2025)

Pattern: ^(0?[1-9]|1[0-2])\/(0?[1-9]|[12]\d|3[01])\/\d{4}$

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: deepseek/deepseek-v4-pro

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

modelstring · enumRequiredPossible values:

rolestring · enumRequiredPossible values:

contentany ofRequired

stringOptional

itemsone ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

typestring · enumRequired

The type of the image.

Possible values:

media_typestring · enumRequired

The media type of the image.

Possible values:

datastringRequired

The base64 encoded image data.

typestring · enumRequiredPossible values:

thinkingstringRequired

signaturestringRequired

typestring · enumRequiredPossible values:

tool_use_idstringRequired

is_errorbooleanOptional

contentany ofOptional

stringOptional

itemsone ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

typestring · enumRequired

The type of the image.

Possible values:

media_typestring · enumRequired

The media type of the image.

Possible values:

datastringRequired

The base64 encoded image data.

idstringRequired

Other propertiesany · nullableOptional

namestringRequired

typestring · enumRequiredPossible values:

datastringRequired

typestring · enumRequiredPossible values:

sourceany ofRequired

typestring · enumRequiredPossible values:

media_typestring · enumRequiredPossible values:

datastringRequired

typestring · enumRequiredPossible values:

datastringRequired

Other propertiesstringOptional

stop_sequencesstring[]Optional

Custom text sequences that will cause the model to stop generating.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

systemstringOptional

A system prompt is a way of providing context and instructions to Claude, such as specifying a particular goal or role.

namestringRequired

Name of the tool.

descriptionstringOptional

Description of what this tool does. Tool descriptions should be as detailed as possible. The more information that the model has about what the tool is and how to use it, the better it will perform. You can use natural language descriptions to reinforce important aspects of the tool input JSON schema.

typestring · enumRequiredPossible values:

propertiesany · nullableOptional

Other propertiesany · nullableOptional

typestring · enumRequiredPossible values:

namestringRequired

typestring · enumRequiredPossible values:

budget_tokensinteger · min: 1024Required

Determines how many tokens Claude can use for its internal reasoning process. Larger budgets can enable more thorough analysis for complex problems, improving response quality. Must be ≥1024 and less than max_tokens.

typestring · enumRequiredPossible values:

max_tokensnumberOptional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

Default: 32000

temperaturenumber · max: 1Optional

Amount of randomness injected into the response. Defaults to 1.0. Ranges from 0.0 to 1.0. Use temperature closer to 0.0 for analytical / multiple choice, and closer to 1.0 for creative and generative tasks. Note that even with temperature of 0.0, the results will not be fully deterministic.

top_pnumber · max: 1Optional

top_knumberOptional

Only sample from the top K options for each subsequent token. Used to remove "long tail" low probability responses. Recommended for advanced use cases only. You usually only need to use temperature.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: anthropic/claude-opus-4

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: baidu/ernie-x1-turbo-32k

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

urlstring · uriRequired

Either a URL of the image or the base64 encoded image data.

detailstring · enumOptional

Specifies the detail level of the image. Currently supports JPG/JPEG, PNG, GIF, and WEBP formats.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

Other propertiesnumber · min: -100 · max: 100Optional

logprobsboolean · nullableOptional

Whether to return log probabilities of the output tokens or not. If True, returns the log probabilities of each output token returned in the content of message.

top_logprobsnumber · max: 20 · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

reasoning_effortstring · enumOptional

Possible values:

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: alibaba/qwen3.6-27b

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

urlstring · uriRequired

Either a URL of the image or the base64 encoded image data.

detailstring · enumOptional

Specifies the detail level of the image. Currently supports JPG/JPEG, PNG, GIF, and WEBP formats.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

ninteger · min: 1 · nullableOptional

How many chat completion choices to generate for each input message. Note that you will be charged based on the number of generated tokens across all of the choices. Keep n as 1 to minimize costs.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

reasoning_effortstring · enumOptional

Possible values:

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: google/gemini-3-1-flash-lite-preview

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: baidu/ernie-x1-1-preview

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

modelstring · enumRequiredPossible values:

rolestring · enumRequiredPossible values:

contentany ofRequired

stringOptional

itemsone ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

typestring · enumRequired

The type of the image.

Possible values:

media_typestring · enumRequired

The media type of the image.

Possible values:

datastringRequired

The base64 encoded image data.

typestring · enumRequiredPossible values:

thinkingstringRequired

signaturestringRequired

typestring · enumRequiredPossible values:

tool_use_idstringRequired

is_errorbooleanOptional

contentany ofOptional

stringOptional

itemsone ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

typestring · enumRequired

The type of the image.

Possible values:

media_typestring · enumRequired

The media type of the image.

Possible values:

datastringRequired

The base64 encoded image data.

idstringRequired

Other propertiesany · nullableOptional

namestringRequired

typestring · enumRequiredPossible values:

datastringRequired

typestring · enumRequiredPossible values:

sourceany ofRequired

typestring · enumRequiredPossible values:

media_typestring · enumRequiredPossible values:

datastringRequired

typestring · enumRequiredPossible values:

datastringRequired

Other propertiesstringOptional

stop_sequencesstring[]Optional

Custom text sequences that will cause the model to stop generating.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

systemstringOptional

A system prompt is a way of providing context and instructions to Claude, such as specifying a particular goal or role.

namestringRequired

Name of the tool.

descriptionstringOptional

typestring · enumRequiredPossible values:

propertiesany · nullableOptional

Other propertiesany · nullableOptional

typestring · enumRequiredPossible values:

namestringRequired

typestring · enumRequiredPossible values:

budget_tokensinteger · min: 1024Required

typestring · enumRequiredPossible values:

max_tokensnumberOptional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

Default: 32000

temperaturenumber · max: 1Optional

top_pnumber · max: 1Optional

top_knumberOptional

Only sample from the top K options for each subsequent token. Used to remove "long tail" low probability responses. Recommended for advanced use cases only. You usually only need to use temperature.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: anthropic/claude-sonnet-4.6

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

contentany ofRequired

The contents of the developer message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

rolestring · enumRequired

The role of the author of the message — in this case, the developer.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

echobooleanOptional

If True, the response will contain the prompt. Can be used with logprobs to return prompt logprobs.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

ninteger · min: 1 · nullableOptional

How many chat completion choices to generate for each input message. Note that you will be charged based on the number of generated tokens across all of the choices. Keep n as 1 to minimize costs.

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

logprobsboolean · nullableOptional

Whether to return log probabilities of the output tokens or not. If True, returns the log probabilities of each output token returned in the content of message.

top_logprobsnumber · max: 20 · nullableOptional

Other propertiesnumber · min: -100 · max: 100Optional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

min_pnumber · min: 0.001 · max: 0.999Optional

A number between 0.001 and 0.999 that can be used as an alternative to top_p and top_k.

top_knumberOptional

Only sample from the top K options for each subsequent token. Used to remove "long tail" low probability responses. Recommended for advanced use cases only. You usually only need to use temperature.

repetition_penaltynumber · nullableOptional

A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: Qwen/Qwen2.5-7B-Instruct-Turbo

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequiredPossible values:

contentany ofRequired

stringOptional

itemsone ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

typestring · enumRequired

The type of the image.

Possible values:

media_typestring · enumRequired

The media type of the image.

Possible values:

datastringRequired

The base64 encoded image data.

typestring · enumRequiredPossible values:

thinkingstringRequired

signaturestringRequired

typestring · enumRequiredPossible values:

tool_use_idstringRequired

is_errorbooleanOptional

contentany ofOptional

stringOptional

itemsone ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

typestring · enumRequired

The type of the image.

Possible values:

media_typestring · enumRequired

The media type of the image.

Possible values:

datastringRequired

The base64 encoded image data.

idstringRequired

Other propertiesany · nullableOptional

namestringRequired

typestring · enumRequiredPossible values:

datastringRequired

typestring · enumRequiredPossible values:

sourceany ofRequired

typestring · enumRequiredPossible values:

media_typestring · enumRequiredPossible values:

datastringRequired

typestring · enumRequiredPossible values:

datastringRequired

Other propertiesstringOptional

stop_sequencesstring[]Optional

Custom text sequences that will cause the model to stop generating.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

systemstringOptional

A system prompt is a way of providing context and instructions to Claude, such as specifying a particular goal or role.

namestringRequired

Name of the tool.

descriptionstringOptional

typestring · enumRequiredPossible values:

propertiesany · nullableOptional

Other propertiesany · nullableOptional

typestring · enumRequiredPossible values:

namestringRequired

typestring · enumRequiredPossible values:

budget_tokensinteger · min: 1024Required

typestring · enumRequiredPossible values:

max_tokensnumberOptional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

Default: 32000

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: anthropic/claude-opus-4-7

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: baidu/ernie-4.5-21b-a3b-thinking

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

post

Body

modelstring · enumRequiredPossible values:

messagesany ofRequired

A list of messages comprising the conversation so far. Depending on the model you use, different message types (modalities) are supported, like text, documents (txt, pdf), images, and audio.

rolestring · enumRequiredPossible values:

contentany ofRequired

stringOptional

itemsone ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

sourceone ofRequired

typestring · enumRequired

The type of the image.

Possible values:

media_typestring · enumRequired

The media type of the image.

Possible values:

datastringRequired

The base64 encoded image data.

typestring · enumRequiredPossible values:

urlstring · uriRequired

typestring · enumRequiredPossible values:

thinkingstringRequired

signaturestringRequired

typestring · enumRequiredPossible values:

tool_use_idstringRequired

is_errorbooleanOptional

contentany ofOptional

stringOptional

itemsone ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

sourceone ofRequired

typestring · enumRequired

The type of the image.

Possible values:

media_typestring · enumRequired

The media type of the image.

Possible values:

datastringRequired

The base64 encoded image data.

typestring · enumRequiredPossible values:

urlstring · uriRequired

typestring · enumRequiredPossible values:

sourcestringRequired

titlestringRequired

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

sourceone ofRequired

typestring · enumRequiredPossible values:

media_typestring · enumRequiredPossible values:

datastringRequired

typestring · enumRequiredPossible values:

media_typestring · enumRequiredPossible values:

datastringRequired

typestring · enumRequiredPossible values:

urlstring · uriRequired

typestring · enumRequiredPossible values:

contentany ofRequired

stringOptional

itemsany · nullableOptional

titlestringOptional

contextstringOptional

typestring · enumRequiredPossible values:

tool_namestringRequired

typestring · enumRequiredPossible values:

idstringRequired

namestringRequired

Other propertiesany · nullableOptional

callerone ofOptional

typestring · enumRequiredPossible values:

tool_idstringRequired

typestring · enumRequiredPossible values:

tool_idstringRequired

typestring · enumRequiredPossible values:

idstringRequired

namestring · enumRequiredPossible values:

Other propertiesany · nullableOptional

callerone ofOptional

typestring · enumRequiredPossible values:

tool_idstringRequired

typestring · enumRequiredPossible values:

tool_idstringRequired

typestring · enumRequiredPossible values:

sourcestringRequired

titlestringRequired

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

tool_use_idstringRequired

contentany ofRequired

typestring · enumRequiredPossible values:

titlestringRequired

urlstringRequired

page_agestringOptional

encrypted_contentstringRequired

typestring · enumRequiredPossible values:

error_codestring · enumRequiredPossible values:

callerone ofOptional

typestring · enumRequiredPossible values:

tool_idstringRequired

typestring · enumRequiredPossible values:

tool_idstringRequired

typestring · enumRequiredPossible values:

tool_use_idstringRequired

contentany ofRequired

typestring · enumRequiredPossible values:

error_codestring · enumRequiredPossible values:

typestring · enumRequiredPossible values:

urlstringRequired

retrieved_atstringOptional

typestring · enumRequiredPossible values:

sourceone ofRequired

typestring · enumRequiredPossible values:

media_typestring · enumRequiredPossible values:

datastringRequired

typestring · enumRequiredPossible values:

media_typestring · enumRequiredPossible values:

datastringRequired

typestring · enumRequiredPossible values:

urlstring · uriRequired

typestring · enumRequiredPossible values:

contentany ofRequired

stringOptional

itemsany · nullableOptional

titlestringOptional

contextstringOptional

callerone ofOptional

typestring · enumRequiredPossible values:

tool_idstringRequired

typestring · enumRequiredPossible values:

tool_idstringRequired

typestring · enumRequiredPossible values:

tool_use_idstringRequired

contentone ofRequired

typestring · enumRequiredPossible values:

error_codestring · enumRequiredPossible values:

typestring · enumRequiredPossible values:

stdoutstringRequired

stderrstringRequired

return_codenumberRequired

typestring · enumRequiredPossible values:

file_idstringRequired

typestring · enumRequiredPossible values:

encrypted_stdoutstringRequired

stderrstringRequired

return_codenumberRequired

typestring · enumRequiredPossible values:

file_idstringRequired

typestring · enumRequiredPossible values:

tool_use_idstringRequired

contentany ofRequired

typestring · enumRequiredPossible values:

error_codestring · enumRequiredPossible values:

typestring · enumRequiredPossible values:

stdoutstringRequired

stderrstringRequired

return_codenumberRequired

typestring · enumRequiredPossible values:

file_idstringRequired

typestring · enumRequiredPossible values:

tool_use_idstringRequired

contentone ofRequired

typestring · enumRequiredPossible values:

error_codestring · enumRequiredPossible values:

error_messagestringOptional

typestring · enumRequiredPossible values:

contentstringRequired

file_typestring · enumRequiredPossible values:

start_linenumberOptional

num_linesnumberOptional

total_linesnumberOptional

typestring · enumRequiredPossible values:

is_file_updatebooleanRequired

typestring · enumRequiredPossible values:

old_startnumberOptional

old_linesnumberOptional

new_startnumberOptional

new_linesnumberOptional

linesstring[]Optional

typestring · enumRequiredPossible values:

tool_use_idstringRequired

contentone ofRequired

typestring · enumRequiredPossible values:

error_codestring · enumRequiredPossible values:

typestring · enumRequiredPossible values:

tool_namestringRequired

typestring · enumRequiredPossible values:

datastringRequired

typestring · enumRequiredPossible values:

sourceone ofRequired

typestring · enumRequiredPossible values:

media_typestring · enumRequiredPossible values:

datastringRequired

typestring · enumRequiredPossible values:

media_typestring · enumRequiredPossible values:

datastringRequired

typestring · enumRequiredPossible values:

urlstring · uriRequired

typestring · enumRequiredPossible values:

contentany ofRequired

stringOptional

itemsany · nullableOptional

titlestringOptional

contextstringOptional

itemsone ofOptional

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequiredPossible values:

urlany ofRequired

Either a URL of the image or the base64 encoded image data.

string · uriOptional

stringOptional

detailstring · enumOptional

Specifies the detail level of the image. Currently supports JPG/JPEG, PNG, GIF, and WEBP formats.

Possible values:

typestring · enumRequiredPossible values:

media_typestring · enumRequiredPossible values:

datastringRequired

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequired

The type of the content part.

Possible values:

dataany ofRequired

Either a URL of the audio or the base64 encoded audio data.

string · uriOptional

stringOptional

formatstring · enumRequired

The format of the encoded audio data. Currently supports "wav" and "mp3".

Possible values:

typestring · enumRequired

The type of the content part.

Possible values:

file_datastringOptional

filenamestringOptional

The file name specified by the user. This name can be used to reference the file when interacting with the model, especially if multiple files are uploaded.

typestring · enumRequiredPossible values:

urlany ofRequired

Base64-encoded local video file.

string · uriOptional

stringOptional

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentany ofRequired

The contents of the tool message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequiredPossible values:

contentstringRequired

namestringRequired

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

any · nullableOptional

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the custom tool to call.

inputstringRequired

The input for the custom tool call generated by the model.

refusalstring · nullableOptional

The refusal message by the Assistant.

idstringRequired

Unique identifier for a previous audio response from the model.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

contentany ofRequired

The contents of the developer message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

rolestring · enumRequired

The role of the author of the message — in this case, the developer.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

Other propertiesstringOptional

stop_sequencesstring[]Optional

Custom text sequences that will cause the model to stop generating.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

systemany ofOptional

A system prompt is a way of providing context and instructions to Claude, such as specifying a particular goal or role.

stringOptional

typestring · enumOptionalDefault: textPossible values:

textstringRequired

typestring · enumRequiredPossible values:

cited_textstringRequired

document_indexnumberRequired

document_titlestringRequired

end_char_indexnumberRequired

start_char_indexnumberRequired

typestring · enumRequiredPossible values:

cited_textstringRequired

document_indexnumberRequired

document_titlestringRequired

end_page_numbernumberRequired

start_page_numbernumberRequired

typestring · enumRequiredPossible values:

cited_textstringRequired

document_indexnumberRequired

document_titlestringRequired

end_block_indexnumberRequired

start_block_indexnumberRequired

typestring · enumRequiredPossible values:

cited_textstringRequired

encrypted_indexstringRequired

titlestringRequired

urlstringRequired

typestring · enumRequiredPossible values:

cited_textstringRequired

end_block_indexnumberRequired

search_result_indexnumberRequired

sourcestringRequired

start_block_indexnumberRequired

titlestringRequired

tool_choiceany ofOptional

typestring · enumRequiredPossible values:

disable_parallel_tool_usebooleanOptional

typestring · enumRequiredPossible values:

disable_parallel_tool_usebooleanOptional

namestringRequired

typestring · enumRequiredPossible values:

disable_parallel_tool_usebooleanOptional

typestring · enumRequiredPossible values:

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

modestring · enumRequired

Constrains the tools available to the model to a pre-defined set.

auto allows the model to pick from among the allowed tools and generate a message.
required requires the model to call one or more of the allowed tools.

Possible values:

Other propertiesany · nullableOptional

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

functionstring · enumRequired

The name of the function to call.

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

customstring · enumRequired

The name of the custom tool to call.

Possible values:

toolsany ofOptional

Definitions of tools that the model may use. If you include tools in your API request, the model may return tool_use content blocks that represent the model's use of those tools. You can then run those tools using the tool input generated by the model and then optionally return results back to the model using tool_result content blocks. Each tool definition includes: name: Name of the tool. description: Optional, but strongly-recommended description of the tool. input_schema: JSON schema for the tool input shape that the model will produce in tool_use output content blocks.

itemsone ofOptional

namestringRequired

Name of the tool.

descriptionstringOptional

typestring · enumRequiredPossible values:

propertiesany · nullableOptional

requiredstring[]Optional

Other propertiesany · nullableOptional

typestring · enumOptionalPossible values:

defer_loadingbooleanOptional

eager_input_streamingbooleanOptional

Other propertiesany · nullableOptional

strictbooleanOptional

itemsstring · enumOptionalPossible values:

namestring · enumOptionalDefault: bashPossible values:

typestring · enumRequiredPossible values:

Other propertiesany · nullableOptional

defer_loadingbooleanOptional

strictbooleanOptional

itemsstring · enumOptionalPossible values:

namestring · enumOptionalDefault: code_executionPossible values:

typestring · enumRequiredPossible values:

defer_loadingbooleanOptional

strictbooleanOptional

itemsstring · enumOptionalPossible values:

namestring · enumOptionalDefault: code_executionPossible values:

typestring · enumRequiredPossible values:

defer_loadingbooleanOptional

strictbooleanOptional

itemsstring · enumOptionalPossible values:

namestring · enumOptionalDefault: code_executionPossible values:

typestring · enumRequiredPossible values:

defer_loadingbooleanOptional

strictbooleanOptional

itemsstring · enumOptionalPossible values:

namestring · enumOptionalDefault: memoryPossible values:

typestring · enumRequiredPossible values:

Other propertiesany · nullableOptional

defer_loadingbooleanOptional

strictbooleanOptional

itemsstring · enumOptionalPossible values:

namestring · enumOptionalDefault: str_replace_editorPossible values:

typestring · enumRequiredPossible values:

Other propertiesany · nullableOptional

defer_loadingbooleanOptional

strictbooleanOptional

itemsstring · enumOptionalPossible values:

namestring · enumOptionalDefault: str_replace_based_edit_toolPossible values:

typestring · enumRequiredPossible values:

Other propertiesany · nullableOptional

defer_loadingbooleanOptional

strictbooleanOptional

itemsstring · enumOptionalPossible values:

namestring · enumOptionalDefault: str_replace_based_edit_toolPossible values:

typestring · enumRequiredPossible values:

max_charactersnumberOptional

Other propertiesany · nullableOptional

defer_loadingbooleanOptional

strictbooleanOptional

itemsstring · enumOptionalPossible values:

namestring · enumOptionalDefault: web_searchPossible values:

typestring · enumRequiredPossible values:

allowed_domainsstring[]Optional

blocked_domainsstring[]Optional

max_usesnumberOptional

typestring · enumRequiredPossible values:

citystringOptional

countrystringOptional

regionstringOptional

timezonestringOptional

defer_loadingbooleanOptional

strictbooleanOptional

itemsstring · enumOptionalPossible values:

namestring · enumOptionalDefault: web_searchPossible values:

typestring · enumRequiredPossible values:

allowed_domainsstring[]Optional

blocked_domainsstring[]Optional

max_usesnumberOptional

typestring · enumRequiredPossible values:

citystringOptional

countrystringOptional

regionstringOptional

timezonestringOptional

defer_loadingbooleanOptional

strictbooleanOptional

itemsstring · enumOptionalPossible values:

namestring · enumOptionalDefault: web_fetchPossible values:

typestring · enumRequiredPossible values:

allowed_domainsstring[]Optional

blocked_domainsstring[]Optional

enabledbooleanOptional

max_content_tokensnumberOptional

max_usesnumberOptional

defer_loadingbooleanOptional

strictbooleanOptional

itemsstring · enumOptionalPossible values:

namestring · enumOptionalDefault: web_fetchPossible values:

typestring · enumRequiredPossible values:

allowed_domainsstring[]Optional

blocked_domainsstring[]Optional

enabledbooleanOptional

max_content_tokensnumberOptional

max_usesnumberOptional

defer_loadingbooleanOptional

strictbooleanOptional

itemsstring · enumOptionalPossible values:

namestring · enumOptionalDefault: web_fetchPossible values:

typestring · enumRequiredPossible values:

allowed_domainsstring[]Optional

blocked_domainsstring[]Optional

enabledbooleanOptional

max_content_tokensnumberOptional

max_usesnumberOptional

defer_loadingbooleanOptional

strictbooleanOptional

itemsstring · enumOptionalPossible values:

namestring · enumOptionalDefault: tool_search_tool_bm25Possible values:

typestring · enumRequiredPossible values:

defer_loadingbooleanOptional

strictbooleanOptional

itemsstring · enumOptionalPossible values:

namestring · enumOptionalDefault: tool_search_tool_regexPossible values:

typestring · enumRequiredPossible values:

defer_loadingbooleanOptional

strictbooleanOptional

itemsstring · enumOptionalPossible values:

itemsone ofOptional

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the custom tool, used to identify it in tool calls.

descriptionstringOptional

Optional description of the custom tool, used to provide more context.

formatone ofRequired

The input format for the custom tool. Default is unconstrained text.

typestring · enumRequiredPossible values:

definitionstringRequired

The grammar definition.

syntaxstring · enumRequired

The syntax of the grammar definition.

Possible values:

thinkingone ofOptional

Configuration for enabling Claude's extended thinking. When enabled, responses include thinking content blocks showing Claude's thinking process before the final answer. Requires a minimum budget of 1,024 tokens and counts towards your max_tokens limit.

typestring · enumRequiredPossible values:

budget_tokensinteger · min: 1024Required

displaystring · enumOptionalDefault: summarizedPossible values:

typestring · enumRequiredPossible values:

displaystring · enumOptionalDefault: summarizedPossible values:

max_tokensnumberOptional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

Default: 64000

temperaturenumber · max: 1Optional

top_pnumber · max: 1Optional

top_knumberOptional

Only sample from the top K options for each subsequent token. Used to remove "long tail" low probability responses. Recommended for advanced use cases only. You usually only need to use temperature.

modelstring · enumRequiredPossible values:

messagesany ofRequired

A list of messages comprising the conversation so far. Depending on the model you use, different message types (modalities) are supported, like text, documents (txt, pdf), images, and audio.

rolestring · enumRequiredPossible values:

contentany ofRequired

stringOptional

itemsone ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

sourceone ofRequired

typestring · enumRequired

The type of the image.

Possible values:

media_typestring · enumRequired

The media type of the image.

Possible values:

datastringRequired

The base64 encoded image data.

typestring · enumRequiredPossible values:

urlstring · uriRequired

typestring · enumRequiredPossible values:

thinkingstringRequired

signaturestringRequired

typestring · enumRequiredPossible values:

tool_use_idstringRequired

is_errorbooleanOptional

contentany ofOptional

stringOptional

itemsone ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

sourceone ofRequired

typestring · enumRequired

The type of the image.

Possible values:

media_typestring · enumRequired

The media type of the image.

Possible values:

datastringRequired

The base64 encoded image data.

typestring · enumRequiredPossible values:

urlstring · uriRequired

typestring · enumRequiredPossible values:

sourcestringRequired

titlestringRequired

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

sourceone ofRequired

typestring · enumRequiredPossible values:

media_typestring · enumRequiredPossible values:

datastringRequired

typestring · enumRequiredPossible values:

media_typestring · enumRequiredPossible values:

datastringRequired

typestring · enumRequiredPossible values:

urlstring · uriRequired

typestring · enumRequiredPossible values:

contentany ofRequired

stringOptional

itemsany · nullableOptional

titlestringOptional

contextstringOptional

typestring · enumRequiredPossible values:

tool_namestringRequired

typestring · enumRequiredPossible values:

idstringRequired

namestringRequired

Other propertiesany · nullableOptional

callerone ofOptional

typestring · enumRequiredPossible values:

tool_idstringRequired

typestring · enumRequiredPossible values:

tool_idstringRequired

typestring · enumRequiredPossible values:

idstringRequired

namestring · enumRequiredPossible values:

Other propertiesany · nullableOptional

callerone ofOptional

typestring · enumRequiredPossible values:

tool_idstringRequired

typestring · enumRequiredPossible values:

tool_idstringRequired

typestring · enumRequiredPossible values:

sourcestringRequired

titlestringRequired

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

tool_use_idstringRequired

contentany ofRequired

typestring · enumRequiredPossible values:

titlestringRequired

urlstringRequired

page_agestringOptional

encrypted_contentstringRequired

typestring · enumRequiredPossible values:

error_codestring · enumRequiredPossible values:

callerone ofOptional

typestring · enumRequiredPossible values:

tool_idstringRequired

typestring · enumRequiredPossible values:

tool_idstringRequired

typestring · enumRequiredPossible values:

tool_use_idstringRequired

contentany ofRequired

typestring · enumRequiredPossible values:

error_codestring · enumRequiredPossible values:

typestring · enumRequiredPossible values:

urlstringRequired

retrieved_atstringOptional

typestring · enumRequiredPossible values:

sourceone ofRequired

typestring · enumRequiredPossible values:

media_typestring · enumRequiredPossible values:

datastringRequired

typestring · enumRequiredPossible values:

media_typestring · enumRequiredPossible values:

datastringRequired

typestring · enumRequiredPossible values:

urlstring · uriRequired

typestring · enumRequiredPossible values:

contentany ofRequired

stringOptional

itemsany · nullableOptional

titlestringOptional

contextstringOptional

callerone ofOptional

typestring · enumRequiredPossible values:

tool_idstringRequired

typestring · enumRequiredPossible values:

tool_idstringRequired

typestring · enumRequiredPossible values:

tool_use_idstringRequired

contentone ofRequired

typestring · enumRequiredPossible values:

error_codestring · enumRequiredPossible values:

typestring · enumRequiredPossible values:

stdoutstringRequired

stderrstringRequired

return_codenumberRequired

typestring · enumRequiredPossible values:

file_idstringRequired

typestring · enumRequiredPossible values:

encrypted_stdoutstringRequired

stderrstringRequired

return_codenumberRequired

typestring · enumRequiredPossible values:

file_idstringRequired

typestring · enumRequiredPossible values:

tool_use_idstringRequired

contentany ofRequired

typestring · enumRequiredPossible values:

error_codestring · enumRequiredPossible values:

typestring · enumRequiredPossible values:

stdoutstringRequired

stderrstringRequired

return_codenumberRequired

typestring · enumRequiredPossible values:

file_idstringRequired

typestring · enumRequiredPossible values:

tool_use_idstringRequired

contentone ofRequired

typestring · enumRequiredPossible values:

error_codestring · enumRequiredPossible values:

error_messagestringOptional

typestring · enumRequiredPossible values:

contentstringRequired

file_typestring · enumRequiredPossible values:

start_linenumberOptional

num_linesnumberOptional

total_linesnumberOptional

typestring · enumRequiredPossible values:

is_file_updatebooleanRequired

typestring · enumRequiredPossible values:

old_startnumberOptional

old_linesnumberOptional

new_startnumberOptional

new_linesnumberOptional

linesstring[]Optional

typestring · enumRequiredPossible values:

tool_use_idstringRequired

contentone ofRequired

typestring · enumRequiredPossible values:

error_codestring · enumRequiredPossible values:

typestring · enumRequiredPossible values:

tool_namestringRequired

typestring · enumRequiredPossible values:

datastringRequired

typestring · enumRequiredPossible values:

sourceone ofRequired

typestring · enumRequiredPossible values:

media_typestring · enumRequiredPossible values:

datastringRequired

typestring · enumRequiredPossible values:

media_typestring · enumRequiredPossible values:

datastringRequired

typestring · enumRequiredPossible values:

urlstring · uriRequired

typestring · enumRequiredPossible values:

contentany ofRequired

stringOptional

itemsany · nullableOptional

titlestringOptional

contextstringOptional

itemsone ofOptional

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequiredPossible values:

urlany ofRequired

Either a URL of the image or the base64 encoded image data.

string · uriOptional

stringOptional

detailstring · enumOptional

Specifies the detail level of the image. Currently supports JPG/JPEG, PNG, GIF, and WEBP formats.

Possible values:

typestring · enumRequiredPossible values:

media_typestring · enumRequiredPossible values:

datastringRequired

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequired

The type of the content part.

Possible values:

dataany ofRequired

Either a URL of the audio or the base64 encoded audio data.

string · uriOptional

stringOptional

formatstring · enumRequired

The format of the encoded audio data. Currently supports "wav" and "mp3".

Possible values:

typestring · enumRequired

The type of the content part.

Possible values:

file_datastringOptional

filenamestringOptional

The file name specified by the user. This name can be used to reference the file when interacting with the model, especially if multiple files are uploaded.

typestring · enumRequiredPossible values:

urlany ofRequired

Base64-encoded local video file.

string · uriOptional

stringOptional

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentany ofRequired

The contents of the tool message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequiredPossible values:

contentstringRequired

namestringRequired

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

any · nullableOptional

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the custom tool to call.

inputstringRequired

The input for the custom tool call generated by the model.

refusalstring · nullableOptional

The refusal message by the Assistant.

idstringRequired

Unique identifier for a previous audio response from the model.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

contentany ofRequired

The contents of the developer message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

rolestring · enumRequired

The role of the author of the message — in this case, the developer.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

Other propertiesstringOptional

stop_sequencesstring[]Optional

Custom text sequences that will cause the model to stop generating.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

systemany ofOptional

A system prompt is a way of providing context and instructions to Claude, such as specifying a particular goal or role.

stringOptional

typestring · enumOptionalDefault: textPossible values:

textstringRequired

typestring · enumRequiredPossible values:

cited_textstringRequired

document_indexnumberRequired

document_titlestringRequired

end_char_indexnumberRequired

start_char_indexnumberRequired

typestring · enumRequiredPossible values:

cited_textstringRequired

document_indexnumberRequired

document_titlestringRequired

end_page_numbernumberRequired

start_page_numbernumberRequired

typestring · enumRequiredPossible values:

cited_textstringRequired

document_indexnumberRequired

document_titlestringRequired

end_block_indexnumberRequired

start_block_indexnumberRequired

typestring · enumRequiredPossible values:

cited_textstringRequired

encrypted_indexstringRequired

titlestringRequired

urlstringRequired

typestring · enumRequiredPossible values:

cited_textstringRequired

end_block_indexnumberRequired

search_result_indexnumberRequired

sourcestringRequired

start_block_indexnumberRequired

titlestringRequired

tool_choiceany ofOptional

typestring · enumRequiredPossible values:

disable_parallel_tool_usebooleanOptional

typestring · enumRequiredPossible values:

disable_parallel_tool_usebooleanOptional

namestringRequired

typestring · enumRequiredPossible values:

disable_parallel_tool_usebooleanOptional

typestring · enumRequiredPossible values:

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

modestring · enumRequired

Constrains the tools available to the model to a pre-defined set.

auto allows the model to pick from among the allowed tools and generate a message.
required requires the model to call one or more of the allowed tools.

Possible values:

Other propertiesany · nullableOptional

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

functionstring · enumRequired

The name of the function to call.

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

customstring · enumRequired

The name of the custom tool to call.

Possible values:

toolsany ofOptional

itemsone ofOptional

namestringRequired

Name of the tool.

descriptionstringOptional

typestring · enumRequiredPossible values:

propertiesany · nullableOptional

requiredstring[]Optional

Other propertiesany · nullableOptional

typestring · enumOptionalPossible values:

defer_loadingbooleanOptional

eager_input_streamingbooleanOptional

Other propertiesany · nullableOptional

strictbooleanOptional

itemsstring · enumOptionalPossible values:

namestring · enumOptionalDefault: bashPossible values:

typestring · enumRequiredPossible values:

Other propertiesany · nullableOptional

defer_loadingbooleanOptional

strictbooleanOptional

itemsstring · enumOptionalPossible values:

namestring · enumOptionalDefault: code_executionPossible values:

typestring · enumRequiredPossible values:

defer_loadingbooleanOptional

strictbooleanOptional

itemsstring · enumOptionalPossible values:

namestring · enumOptionalDefault: code_executionPossible values:

typestring · enumRequiredPossible values:

defer_loadingbooleanOptional

strictbooleanOptional

itemsstring · enumOptionalPossible values:

namestring · enumOptionalDefault: code_executionPossible values:

typestring · enumRequiredPossible values:

defer_loadingbooleanOptional

strictbooleanOptional

itemsstring · enumOptionalPossible values:

namestring · enumOptionalDefault: memoryPossible values:

typestring · enumRequiredPossible values:

Other propertiesany · nullableOptional

defer_loadingbooleanOptional

strictbooleanOptional

itemsstring · enumOptionalPossible values:

namestring · enumOptionalDefault: str_replace_editorPossible values:

typestring · enumRequiredPossible values:

Other propertiesany · nullableOptional

defer_loadingbooleanOptional

strictbooleanOptional

itemsstring · enumOptionalPossible values:

namestring · enumOptionalDefault: str_replace_based_edit_toolPossible values:

typestring · enumRequiredPossible values:

Other propertiesany · nullableOptional

defer_loadingbooleanOptional

strictbooleanOptional

itemsstring · enumOptionalPossible values:

namestring · enumOptionalDefault: str_replace_based_edit_toolPossible values:

typestring · enumRequiredPossible values:

max_charactersnumberOptional

Other propertiesany · nullableOptional

defer_loadingbooleanOptional

strictbooleanOptional

itemsstring · enumOptionalPossible values:

namestring · enumOptionalDefault: web_searchPossible values:

typestring · enumRequiredPossible values:

allowed_domainsstring[]Optional

blocked_domainsstring[]Optional

max_usesnumberOptional

typestring · enumRequiredPossible values:

citystringOptional

countrystringOptional

regionstringOptional

timezonestringOptional

defer_loadingbooleanOptional

strictbooleanOptional

itemsstring · enumOptionalPossible values:

namestring · enumOptionalDefault: web_searchPossible values:

typestring · enumRequiredPossible values:

allowed_domainsstring[]Optional

blocked_domainsstring[]Optional

max_usesnumberOptional

typestring · enumRequiredPossible values:

citystringOptional

countrystringOptional

regionstringOptional

timezonestringOptional

defer_loadingbooleanOptional

strictbooleanOptional

itemsstring · enumOptionalPossible values:

namestring · enumOptionalDefault: web_fetchPossible values:

typestring · enumRequiredPossible values:

allowed_domainsstring[]Optional

blocked_domainsstring[]Optional

enabledbooleanOptional

max_content_tokensnumberOptional

max_usesnumberOptional

defer_loadingbooleanOptional

strictbooleanOptional

itemsstring · enumOptionalPossible values:

namestring · enumOptionalDefault: web_fetchPossible values:

typestring · enumRequiredPossible values:

allowed_domainsstring[]Optional

blocked_domainsstring[]Optional

enabledbooleanOptional

max_content_tokensnumberOptional

max_usesnumberOptional

defer_loadingbooleanOptional

strictbooleanOptional

itemsstring · enumOptionalPossible values:

namestring · enumOptionalDefault: web_fetchPossible values:

typestring · enumRequiredPossible values:

allowed_domainsstring[]Optional

blocked_domainsstring[]Optional

enabledbooleanOptional

max_content_tokensnumberOptional

max_usesnumberOptional

defer_loadingbooleanOptional

strictbooleanOptional

itemsstring · enumOptionalPossible values:

namestring · enumOptionalDefault: tool_search_tool_bm25Possible values:

typestring · enumRequiredPossible values:

defer_loadingbooleanOptional

strictbooleanOptional

itemsstring · enumOptionalPossible values:

namestring · enumOptionalDefault: tool_search_tool_regexPossible values:

typestring · enumRequiredPossible values:

defer_loadingbooleanOptional

strictbooleanOptional

itemsstring · enumOptionalPossible values:

itemsone ofOptional

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the custom tool, used to identify it in tool calls.

descriptionstringOptional

Optional description of the custom tool, used to provide more context.

formatone ofRequired

The input format for the custom tool. Default is unconstrained text.

typestring · enumRequiredPossible values:

definitionstringRequired

The grammar definition.

syntaxstring · enumRequired

The syntax of the grammar definition.

Possible values:

thinkingone ofOptional

typestring · enumRequiredPossible values:

budget_tokensinteger · min: 1024Required

displaystring · enumOptionalDefault: summarizedPossible values:

typestring · enumRequiredPossible values:

displaystring · enumOptionalDefault: summarizedPossible values:

max_tokensnumberOptional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

Default: 128000

Responses

200Success

application/json

modelstringRequired

Model used to generate the response.

Example: claude-sonnet-4-20250514

idstringRequired

Unique identifier of the generated message.

Example: msg_01SUmNmSRFZsoa6h96MxJEHH

typestringRequired

Object type.

Example: message

rolestringRequired

Role of the message author.

Example: assistant

typestringRequired

Content block type.

Example: text

textstringRequired

Generated text content.

Example:

I'm functioning well, thank you for asking! I'm ready to help you with any questions or tasks you might have. How can I assist you today?

stop_reasonstring · nullableOptional

Reason why generation stopped.

Example: end_turn

stop_sequencestring · nullableOptional

Custom stop sequence that caused generation to stop.

stop_detailsany · nullableOptional

Additional details about why generation stopped.

input_tokensintegerRequired

Number of input tokens used.

Example: 27

cache_creation_input_tokensintegerRequired

Number of tokens used for cache creation.

Example: 0

cache_read_input_tokensintegerRequired

Number of tokens retrieved from cache.

Example: 0

ephemeral_5m_input_tokensintegerRequired

Number of tokens cached for 5 minutes.

Example: 0

ephemeral_1h_input_tokensintegerRequired

Number of tokens cached for 1 hour.

Example: 0

output_tokensintegerRequired

Number of output tokens generated.

Example: 35

service_tierstringRequired

Service tier used for inference.

Example: standard

inference_geostringRequired

Geographic inference location.

Example: not_available

credits_usednumberRequired

Number of credits consumed.

Example: 1576

usd_spentnumberRequired

Amount spent in USD.

Example: 0.000788

post

/v1/messages

200Success

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: baidu/ernie-4.5-21b-a3b

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

urlstring · uriRequired

Either a URL of the image or the base64 encoded image data.

detailstring · enumOptional

Specifies the detail level of the image. Currently supports JPG/JPEG, PNG, GIF, and WEBP formats.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

Other propertiesnumber · min: -100 · max: 100Optional

logprobsboolean · nullableOptional

Whether to return log probabilities of the output tokens or not. If True, returns the log probabilities of each output token returned in the content of message.

top_logprobsnumber · max: 20 · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

reasoning_effortstring · enumOptional

Possible values:

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: alibaba/qwen3.6-35b-a3b

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

urlany ofRequired

Either a URL of the image or the base64 encoded image data.

string · uriOptional

stringOptional

detailstring · enumOptional

Specifies the detail level of the image. Currently supports JPG/JPEG, PNG, GIF, and WEBP formats.

Possible values:

typestring · enumRequiredPossible values:

urlany ofRequired

Base64-encoded local video file.

string · uriOptional

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

file_datastringOptional

filenamestringOptional

The file name specified by the user. This name can be used to reference the file when interacting with the model, especially if multiple files are uploaded.

typestring · enumRequired

The type of the content part.

Possible values:

dataany ofRequired

Either a URL of the audio or the base64 encoded audio data.

string · uriOptional

stringOptional

formatstring · enumRequired

The format of the encoded audio data. Currently supports "wav" and "mp3".

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentany ofRequired

The contents of the tool message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

any · nullableOptional

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the custom tool to call.

inputstringRequired

The input for the custom tool call generated by the model.

refusalstring · nullableOptional

The refusal message by the Assistant.

idstringRequired

Unique identifier for a previous audio response from the model.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

formatstring · enumRequired

Specifies the output audio format. Must be one of wav, mp3, flac, opus, or pcm16.

Possible values:

voiceany ofRequired

The voice the model uses to respond. Supported voices are alloy, ash, ballad, coral, echo, fable, nova, onyx, sage, and shimmer.

string · enumOptionalPossible values:

stringOptional

itemsstring · enumOptionalPossible values:

ninteger · min: 1 · nullableOptional

How many chat completion choices to generate for each input message. Note that you will be charged based on the number of generated tokens across all of the choices. Keep n as 1 to minimize costs.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the custom tool, used to identify it in tool calls.

descriptionstringOptional

Optional description of the custom tool, used to provide more context.

formatone ofRequired

The input format for the custom tool. Default is unconstrained text.

typestring · enumRequiredPossible values:

definitionstringRequired

The grammar definition.

syntaxstring · enumRequired

The syntax of the grammar definition.

Possible values:

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

modestring · enumRequired

Constrains the tools available to the model to a pre-defined set.

auto allows the model to pick from among the allowed tools and generate a message.
required requires the model to call one or more of the allowed tools.

Possible values:

Other propertiesany · nullableOptional

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

functionstring · enumRequired

The name of the function to call.

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

customstring · enumRequired

The name of the custom tool to call.

Possible values:

normalize_tool_schemasbooleanOptional

Enable provider compatibility normalization for tool function JSON schemas.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

reasoning_effortstring · enumOptional

Possible values:

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: google/gemini-3-5-flash

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequired

The type of the content part.

Possible values:

file_datastringOptional

filenamestringOptional

The file name specified by the user. This name can be used to reference the file when interacting with the model, especially if multiple files are uploaded.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

contentany ofRequired

The contents of the developer message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

rolestring · enumRequired

The role of the author of the message — in this case, the developer.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

echobooleanOptional

If True, the response will contain the prompt. Can be used with logprobs to return prompt logprobs.

min_pnumber · min: 0.001 · max: 0.999Optional

A number between 0.001 and 0.999 that can be used as an alternative to top_p and top_k.

top_knumberOptional

Only sample from the top K options for each subsequent token. Used to remove "long tail" low probability responses. Recommended for advanced use cases only. You usually only need to use temperature.

repetition_penaltynumber · nullableOptional

A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.

Other propertiesnumber · min: -100 · max: 100Optional

ninteger · min: 1 · nullableOptional

How many chat completion choices to generate for each input message. Note that you will be charged based on the number of generated tokens across all of the choices. Keep n as 1 to minimize costs.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: deepseek/deepseek-r1

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

Other propertiesnumber · min: -100 · max: 100Optional

logprobsboolean · nullableOptional

Whether to return log probabilities of the output tokens or not. If True, returns the log probabilities of each output token returned in the content of message.

top_logprobsnumber · max: 20 · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

reasoning_effortstring · enumOptional

Possible values:

effortstring · enumOptional

Reasoning effort setting

Possible values:

max_tokensinteger · min: 1Optional

Max tokens of reasoning content. Cannot be used simultaneously with effort.

excludebooleanOptional

Whether to exclude reasoning from the response

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

echobooleanOptional

If True, the response will contain the prompt. Can be used with logprobs to return prompt logprobs.

min_pnumber · min: 0.001 · max: 0.999Optional

A number between 0.001 and 0.999 that can be used as an alternative to top_p and top_k.

top_knumberOptional

Only sample from the top K options for each subsequent token. Used to remove "long tail" low probability responses. Recommended for advanced use cases only. You usually only need to use temperature.

top_anumber · max: 1Optional

Alternate top sampling parameter.

repetition_penaltynumber · nullableOptional

A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.

search_context_sizestring · enumOptional

High level guidance for the amount of context window space to use for the search. One of low, medium, or high. medium is the default.

Possible values:

citystringOptional

Free text input for the city of the user, e.g. San Francisco.

countrystringOptional

The two-letter ISO country code of the user, e.g. US.

Pattern: ^[A-Z]{2}$

regionstringOptional

Free text input for the region of the user, e.g. California.

timezonestringOptional

The IANA timezone of the user, e.g. America/Los_Angeles.

typestring · enumRequired

The type of location approximation. Always approximate.

Possible values:

search_modestring · enumOptional

Controls the search mode used for the request. When set to 'academic', results will prioritize scholarly sources like peer-reviewed papers and academic journals.

Default: academicPossible values:

search_domain_filterstring[]Optional

A list of domains to limit search results to. Currently limited to 10 domains for Allowlisting and Denylisting. For Denylisting, add a - at the beginning of the domain string.

return_imagesbooleanOptional

Determines whether search results should include images.

Default: false

return_related_questionsbooleanOptional

Determines whether related questions should be returned.

Default: false

search_recency_filterstring · enumOptional

Filters search results based on time (e.g., 'week', 'day').

Possible values:

search_after_date_filterstringOptional

Filters search results to only include content published after this date. Format should be %m/%d/%Y (e.g. 3/1/2025)

Pattern: ^(0?[1-9]|1[0-2])\/(0?[1-9]|[12]\d|3[01])\/\d{4}$

search_before_date_filterstringOptional

Filters search results to only include content published before this date. Format should be %m/%d/%Y (e.g. 3/1/2025)

Pattern: ^(0?[1-9]|1[0-2])\/(0?[1-9]|[12]\d|3[01])\/\d{4}$

last_updated_after_filterstringOptional

Filters search results to only include content last updated after this date. Format should be %m/%d/%Y (e.g. 3/1/2025)

Pattern: ^(0?[1-9]|1[0-2])\/(0?[1-9]|[12]\d|3[01])\/\d{4}$

last_updated_before_filterstringOptional

Filters search results to only include content last updated before this date. Format should be %m/%d/%Y (e.g. 3/1/2025)

Pattern: ^(0?[1-9]|1[0-2])\/(0?[1-9]|[12]\d|3[01])\/\d{4}$

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: deepseek/deepseek-v4-flash

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: baidu/ernie-5-0-thinking-preview

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

urlstring · uriRequired

Either a URL of the image or the base64 encoded image data.

detailstring · enumOptional

Specifies the detail level of the image. Currently supports JPG/JPEG, PNG, GIF, and WEBP formats.

Possible values:

typestring · enumRequired

The type of the content part.

Possible values:

file_datastringOptional

filenamestringOptional

The file name specified by the user. This name can be used to reference the file when interacting with the model, especially if multiple files are uploaded.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: baidu/ernie-4.5-vl-28b-a3b

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequiredPossible values:

contentany ofRequired

stringOptional

itemsone ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

typestring · enumRequired

The type of the image.

Possible values:

media_typestring · enumRequired

The media type of the image.

Possible values:

datastringRequired

The base64 encoded image data.

typestring · enumRequiredPossible values:

thinkingstringRequired

signaturestringRequired

typestring · enumRequiredPossible values:

tool_use_idstringRequired

is_errorbooleanOptional

contentany ofOptional

stringOptional

itemsone ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

typestring · enumRequired

The type of the image.

Possible values:

media_typestring · enumRequired

The media type of the image.

Possible values:

datastringRequired

The base64 encoded image data.

idstringRequired

Other propertiesany · nullableOptional

namestringRequired

typestring · enumRequiredPossible values:

datastringRequired

typestring · enumRequiredPossible values:

sourceany ofRequired

typestring · enumRequiredPossible values:

media_typestring · enumRequiredPossible values:

datastringRequired

typestring · enumRequiredPossible values:

datastringRequired

Other propertiesstringOptional

stop_sequencesstring[]Optional

Custom text sequences that will cause the model to stop generating.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

systemstringOptional

A system prompt is a way of providing context and instructions to Claude, such as specifying a particular goal or role.

namestringRequired

Name of the tool.

descriptionstringOptional

typestring · enumRequiredPossible values:

propertiesany · nullableOptional

Other propertiesany · nullableOptional

typestring · enumRequiredPossible values:

namestringRequired

typestring · enumRequiredPossible values:

budget_tokensinteger · min: 1024Required

typestring · enumRequiredPossible values:

max_tokensnumberOptional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

Default: 32000

temperaturenumber · max: 1Optional

top_pnumber · max: 1Optional

top_knumberOptional

Only sample from the top K options for each subsequent token. Used to remove "long tail" low probability responses. Recommended for advanced use cases only. You usually only need to use temperature.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: anthropic/claude-opus-4-5

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequired

The type of the content part.

Possible values:

file_datastringOptional

filenamestringOptional

The file name specified by the user. This name can be used to reference the file when interacting with the model, especially if multiple files are uploaded.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

contentany ofRequired

The contents of the developer message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

rolestring · enumRequired

The role of the author of the message — in this case, the developer.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

min_pnumber · min: 0.001 · max: 0.999Optional

A number between 0.001 and 0.999 that can be used as an alternative to top_p and top_k.

top_knumberOptional

Only sample from the top K options for each subsequent token. Used to remove "long tail" low probability responses. Recommended for advanced use cases only. You usually only need to use temperature.

repetition_penaltynumber · nullableOptional

A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.

top_anumber · max: 1Optional

Alternate top sampling parameter.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: cohere/command-a

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: baidu/ernie-4-5-8k-preview

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

urlany ofRequired

Either a URL of the image or the base64 encoded image data.

string · uriOptional

stringOptional

detailstring · enumOptional

Specifies the detail level of the image. Currently supports JPG/JPEG, PNG, GIF, and WEBP formats.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentany ofRequired

The contents of the tool message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

any · nullableOptional

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the custom tool to call.

inputstringRequired

The input for the custom tool call generated by the model.

refusalstring · nullableOptional

The refusal message by the Assistant.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the custom tool, used to identify it in tool calls.

descriptionstringOptional

Optional description of the custom tool, used to provide more context.

formatone ofRequired

The input format for the custom tool. Default is unconstrained text.

typestring · enumRequiredPossible values:

definitionstringRequired

The grammar definition.

syntaxstring · enumRequired

The syntax of the grammar definition.

Possible values:

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

modestring · enumRequired

Constrains the tools available to the model to a pre-defined set.

auto allows the model to pick from among the allowed tools and generate a message.
required requires the model to call one or more of the allowed tools.

Possible values:

Other propertiesany · nullableOptional

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

functionstring · enumRequired

The name of the function to call.

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

customstring · enumRequired

The name of the custom tool to call.

Possible values:

normalize_tool_schemasbooleanOptional

Enable provider compatibility normalization for tool function JSON schemas.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

Other propertiesnumber · min: -100 · max: 100Optional

logprobsboolean · nullableOptional

Whether to return log probabilities of the output tokens or not. If True, returns the log probabilities of each output token returned in the content of message.

top_logprobsnumber · max: 20 · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

reasoning_effortstring · enumOptional

Possible values:

effortstring · enumOptional

Reasoning effort setting

Possible values:

max_tokensinteger · min: 1Optional

Max tokens of reasoning content. Cannot be used simultaneously with effort.

excludebooleanOptional

Whether to exclude reasoning from the response

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

echobooleanOptional

If True, the response will contain the prompt. Can be used with logprobs to return prompt logprobs.

min_pnumber · min: 0.001 · max: 0.999Optional

A number between 0.001 and 0.999 that can be used as an alternative to top_p and top_k.

top_knumberOptional

Only sample from the top K options for each subsequent token. Used to remove "long tail" low probability responses. Recommended for advanced use cases only. You usually only need to use temperature.

top_anumber · max: 1Optional

Alternate top sampling parameter.

repetition_penaltynumber · nullableOptional

A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.

search_context_sizestring · enumOptional

High level guidance for the amount of context window space to use for the search. One of low, medium, or high. medium is the default.

Possible values:

citystringOptional

Free text input for the city of the user, e.g. San Francisco.

countrystringOptional

The two-letter ISO country code of the user, e.g. US.

Pattern: ^[A-Z]{2}$

regionstringOptional

Free text input for the region of the user, e.g. California.

timezonestringOptional

The IANA timezone of the user, e.g. America/Los_Angeles.

typestring · enumRequired

The type of location approximation. Always approximate.

Possible values:

search_modestring · enumOptional

Controls the search mode used for the request. When set to 'academic', results will prioritize scholarly sources like peer-reviewed papers and academic journals.

Default: academicPossible values:

search_domain_filterstring[]Optional

A list of domains to limit search results to. Currently limited to 10 domains for Allowlisting and Denylisting. For Denylisting, add a - at the beginning of the domain string.

return_imagesbooleanOptional

Determines whether search results should include images.

Default: false

return_related_questionsbooleanOptional

Determines whether related questions should be returned.

Default: false

search_recency_filterstring · enumOptional

Filters search results based on time (e.g., 'week', 'day').

Possible values:

search_after_date_filterstringOptional

Filters search results to only include content published after this date. Format should be %m/%d/%Y (e.g. 3/1/2025)

Pattern: ^(0?[1-9]|1[0-2])\/(0?[1-9]|[12]\d|3[01])\/\d{4}$

search_before_date_filterstringOptional

Filters search results to only include content published before this date. Format should be %m/%d/%Y (e.g. 3/1/2025)

Pattern: ^(0?[1-9]|1[0-2])\/(0?[1-9]|[12]\d|3[01])\/\d{4}$

last_updated_after_filterstringOptional

Filters search results to only include content last updated after this date. Format should be %m/%d/%Y (e.g. 3/1/2025)

Pattern: ^(0?[1-9]|1[0-2])\/(0?[1-9]|[12]\d|3[01])\/\d{4}$

last_updated_before_filterstringOptional

Filters search results to only include content last updated before this date. Format should be %m/%d/%Y (e.g. 3/1/2025)

Pattern: ^(0?[1-9]|1[0-2])\/(0?[1-9]|[12]\d|3[01])\/\d{4}$

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: gpt-4o-2024-08-06

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequired

The type of the content part.

Possible values:

file_datastringOptional

filenamestringOptional

The file name specified by the user. This name can be used to reference the file when interacting with the model, especially if multiple files are uploaded.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

contentany ofRequired

The contents of the developer message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

rolestring · enumRequired

The role of the author of the message — in this case, the developer.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

echobooleanOptional

If True, the response will contain the prompt. Can be used with logprobs to return prompt logprobs.

min_pnumber · min: 0.001 · max: 0.999Optional

A number between 0.001 and 0.999 that can be used as an alternative to top_p and top_k.

top_knumberOptional

Only sample from the top K options for each subsequent token. Used to remove "long tail" low probability responses. Recommended for advanced use cases only. You usually only need to use temperature.

repetition_penaltynumber · nullableOptional

A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.

Other propertiesnumber · min: -100 · max: 100Optional

ninteger · min: 1 · nullableOptional

How many chat completion choices to generate for each input message. Note that you will be charged based on the number of generated tokens across all of the choices. Keep n as 1 to minimize costs.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: deepseek/deepseek-reasoner-v3.1

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequired

The type of the content part.

Possible values:

file_datastringOptional

filenamestringOptional

The file name specified by the user. This name can be used to reference the file when interacting with the model, especially if multiple files are uploaded.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

contentany ofRequired

The contents of the developer message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

rolestring · enumRequired

The role of the author of the message — in this case, the developer.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

echobooleanOptional

If True, the response will contain the prompt. Can be used with logprobs to return prompt logprobs.

min_pnumber · min: 0.001 · max: 0.999Optional

A number between 0.001 and 0.999 that can be used as an alternative to top_p and top_k.

top_knumberOptional

Only sample from the top K options for each subsequent token. Used to remove "long tail" low probability responses. Recommended for advanced use cases only. You usually only need to use temperature.

repetition_penaltynumber · nullableOptional

A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.

Other propertiesnumber · min: -100 · max: 100Optional

top_anumber · max: 1Optional

Alternate top sampling parameter.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: deepseek/deepseek-non-reasoner-v3.1-terminus

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequiredPossible values:

contentany ofRequired

stringOptional

itemsone ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

typestring · enumRequired

The type of the image.

Possible values:

media_typestring · enumRequired

The media type of the image.

Possible values:

datastringRequired

The base64 encoded image data.

typestring · enumRequiredPossible values:

thinkingstringRequired

signaturestringRequired

typestring · enumRequiredPossible values:

tool_use_idstringRequired

is_errorbooleanOptional

contentany ofOptional

stringOptional

itemsone ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

typestring · enumRequired

The type of the image.

Possible values:

media_typestring · enumRequired

The media type of the image.

Possible values:

datastringRequired

The base64 encoded image data.

idstringRequired

Other propertiesany · nullableOptional

namestringRequired

typestring · enumRequiredPossible values:

datastringRequired

typestring · enumRequiredPossible values:

sourceany ofRequired

typestring · enumRequiredPossible values:

media_typestring · enumRequiredPossible values:

datastringRequired

typestring · enumRequiredPossible values:

datastringRequired

Other propertiesstringOptional

stop_sequencesstring[]Optional

Custom text sequences that will cause the model to stop generating.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

systemstringOptional

A system prompt is a way of providing context and instructions to Claude, such as specifying a particular goal or role.

namestringRequired

Name of the tool.

descriptionstringOptional

typestring · enumRequiredPossible values:

propertiesany · nullableOptional

Other propertiesany · nullableOptional

typestring · enumRequiredPossible values:

namestringRequired

typestring · enumRequiredPossible values:

budget_tokensinteger · min: 1024Required

typestring · enumRequiredPossible values:

max_tokensnumberOptional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

Default: 32000

temperaturenumber · max: 1Optional

top_pnumber · max: 1Optional

top_knumberOptional

Only sample from the top K options for each subsequent token. Used to remove "long tail" low probability responses. Recommended for advanced use cases only. You usually only need to use temperature.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: anthropic/claude-opus-4-6

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: alibaba/qwen3-coder-480b-a35b-instruct

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

urlstring · uriRequired

Either a URL of the image or the base64 encoded image data.

detailstring · enumOptional

Specifies the detail level of the image. Currently supports JPG/JPEG, PNG, GIF, and WEBP formats.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

ninteger · min: 1 · nullableOptional

How many chat completion choices to generate for each input message. Note that you will be charged based on the number of generated tokens across all of the choices. Keep n as 1 to minimize costs.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

reasoning_effortstring · enumOptional

Possible values:

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: google/gemini-2.5-flash-lite-preview

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

modelstring · enumRequiredPossible values:

rolestring · enumRequiredPossible values:

contentany ofRequired

stringOptional

itemsone ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

typestring · enumRequired

The type of the image.

Possible values:

media_typestring · enumRequired

The media type of the image.

Possible values:

datastringRequired

The base64 encoded image data.

typestring · enumRequiredPossible values:

thinkingstringRequired

signaturestringRequired

typestring · enumRequiredPossible values:

tool_use_idstringRequired

is_errorbooleanOptional

contentany ofOptional

stringOptional

itemsone ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

typestring · enumRequired

The type of the image.

Possible values:

media_typestring · enumRequired

The media type of the image.

Possible values:

datastringRequired

The base64 encoded image data.

idstringRequired

Other propertiesany · nullableOptional

namestringRequired

typestring · enumRequiredPossible values:

datastringRequired

typestring · enumRequiredPossible values:

sourceany ofRequired

typestring · enumRequiredPossible values:

media_typestring · enumRequiredPossible values:

datastringRequired

typestring · enumRequiredPossible values:

datastringRequired

Other propertiesstringOptional

stop_sequencesstring[]Optional

Custom text sequences that will cause the model to stop generating.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

systemstringOptional

A system prompt is a way of providing context and instructions to Claude, such as specifying a particular goal or role.

namestringRequired

Name of the tool.

descriptionstringOptional

typestring · enumRequiredPossible values:

propertiesany · nullableOptional

Other propertiesany · nullableOptional

typestring · enumRequiredPossible values:

namestringRequired

typestring · enumRequiredPossible values:

budget_tokensinteger · min: 1024Required

typestring · enumRequiredPossible values:

max_tokensnumberOptional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

Default: 32000

temperaturenumber · max: 1Optional

top_pnumber · max: 1Optional

top_knumberOptional

Only sample from the top K options for each subsequent token. Used to remove "long tail" low probability responses. Recommended for advanced use cases only. You usually only need to use temperature.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: anthropic/claude-sonnet-4.5

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: baidu/ernie-4.5-0.3b

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

urlstring · uriRequired

Either a URL of the image or the base64 encoded image data.

detailstring · enumOptional

Specifies the detail level of the image. Currently supports JPG/JPEG, PNG, GIF, and WEBP formats.

Possible values:

typestring · enumRequired

The type of the content part.

Possible values:

urlstring · uriRequired

Either a URL of the video or the base64 encoded video data.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

reasoning_effortstring · enumOptional

Possible values:

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: bytedance/dola-seed-2-0-lite

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentany ofRequired

The contents of the tool message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

any · nullableOptional

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the custom tool to call.

inputstringRequired

The input for the custom tool call generated by the model.

refusalstring · nullableOptional

The refusal message by the Assistant.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the custom tool, used to identify it in tool calls.

descriptionstringOptional

Optional description of the custom tool, used to provide more context.

formatone ofRequired

The input format for the custom tool. Default is unconstrained text.

typestring · enumRequiredPossible values:

definitionstringRequired

The grammar definition.

syntaxstring · enumRequired

The syntax of the grammar definition.

Possible values:

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

modestring · enumRequired

Constrains the tools available to the model to a pre-defined set.

auto allows the model to pick from among the allowed tools and generate a message.
required requires the model to call one or more of the allowed tools.

Possible values:

Other propertiesany · nullableOptional

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

functionstring · enumRequired

The name of the function to call.

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

customstring · enumRequired

The name of the custom tool to call.

Possible values:

normalize_tool_schemasbooleanOptional

Enable provider compatibility normalization for tool function JSON schemas.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

Other propertiesnumber · min: -100 · max: 100Optional

logprobsboolean · nullableOptional

Whether to return log probabilities of the output tokens or not. If True, returns the log probabilities of each output token returned in the content of message.

top_logprobsnumber · max: 20 · nullableOptional

reasoning_effortstring · enumOptional

Possible values:

effortstring · enumOptional

Reasoning effort setting

Possible values:

max_tokensinteger · min: 1Optional

Max tokens of reasoning content. Cannot be used simultaneously with effort.

excludebooleanOptional

Whether to exclude reasoning from the response

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: alibaba/qwen3.7-max

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

urlstring · uriRequired

Either a URL of the image or the base64 encoded image data.

detailstring · enumOptional

Specifies the detail level of the image. Currently supports JPG/JPEG, PNG, GIF, and WEBP formats.

Possible values:

typestring · enumRequired

The type of the content part.

Possible values:

dataany ofRequired

Either a URL of the audio or the base64 encoded audio data.

string · uriOptional

stringOptional

formatstring · enumRequired

The format of the encoded audio data. Currently supports "wav" and "mp3".

Possible values:

typestring · enumRequired

The type of the content part.

Possible values:

urlstring · uriRequired

Either a URL of the video or the base64 encoded video data.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

idstringRequired

Unique identifier for a previous audio response from the model.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

formatstring · enumRequired

Specifies the output audio format. Must be one of wav, mp3, flac, opus, or pcm16.

Possible values:

voiceany ofRequired

The voice the model uses to respond. Supported voices are alloy, ash, ballad, coral, echo, fable, nova, onyx, sage, and shimmer.

string · enumOptionalPossible values:

stringOptional

itemsstring · enumOptionalPossible values:

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

Other propertiesnumber · min: -100 · max: 100Optional

logprobsboolean · nullableOptional

Whether to return log probabilities of the output tokens or not. If True, returns the log probabilities of each output token returned in the content of message.

top_logprobsnumber · max: 20 · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

enable_thinkingbooleanOptional

Specifies whether to use the thinking mode.

Default: false

thinking_budgetinteger · min: 1Optional

The maximum reasoning length, effective only when enable_thinking is set to true.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: alibaba/qwen3.5-omni-flash

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: baidu/ernie-4-5-turbo-128k

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

urlstring · uriRequired

Either a URL of the image or the base64 encoded image data.

detailstring · enumOptional

Specifies the detail level of the image. Currently supports JPG/JPEG, PNG, GIF, and WEBP formats.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: bytedance/seed-1-8

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

urlstring · uriRequired

Either a URL of the image or the base64 encoded image data.

detailstring · enumOptional

Specifies the detail level of the image. Currently supports JPG/JPEG, PNG, GIF, and WEBP formats.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

repetition_penaltynumber · nullableOptional

A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: alibaba/qwen3-vl-32b-instruct

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequired

The type of the content part.

Possible values:

file_datastringOptional

filenamestringOptional

The file name specified by the user. This name can be used to reference the file when interacting with the model, especially if multiple files are uploaded.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

contentany ofRequired

The contents of the developer message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

rolestring · enumRequired

The role of the author of the message — in this case, the developer.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

echobooleanOptional

If True, the response will contain the prompt. Can be used with logprobs to return prompt logprobs.

min_pnumber · min: 0.001 · max: 0.999Optional

A number between 0.001 and 0.999 that can be used as an alternative to top_p and top_k.

top_knumberOptional

Only sample from the top K options for each subsequent token. Used to remove "long tail" low probability responses. Recommended for advanced use cases only. You usually only need to use temperature.

repetition_penaltynumber · nullableOptional

A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.

Other propertiesnumber · min: -100 · max: 100Optional

top_anumber · max: 1Optional

Alternate top sampling parameter.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: deepseek/deepseek-chat-v3.1

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

logprobsboolean · nullableOptional

Whether to return log probabilities of the output tokens or not. If True, returns the log probabilities of each output token returned in the content of message.

top_logprobsnumber · max: 20 · nullableOptional

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: qwen-max

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

urlstring · uriRequired

Either a URL of the image or the base64 encoded image data.

detailstring · enumOptional

Specifies the detail level of the image. Currently supports JPG/JPEG, PNG, GIF, and WEBP formats.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

ninteger · min: 1 · nullableOptional

How many chat completion choices to generate for each input message. Note that you will be charged based on the number of generated tokens across all of the choices. Keep n as 1 to minimize costs.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: google/gemini-3-1-pro-preview

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

contentany ofRequired

The contents of the developer message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

rolestring · enumRequired

The role of the author of the message — in this case, the developer.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

echobooleanOptional

If True, the response will contain the prompt. Can be used with logprobs to return prompt logprobs.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

ninteger · min: 1 · nullableOptional

How many chat completion choices to generate for each input message. Note that you will be charged based on the number of generated tokens across all of the choices. Keep n as 1 to minimize costs.

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

logprobsboolean · nullableOptional

Whether to return log probabilities of the output tokens or not. If True, returns the log probabilities of each output token returned in the content of message.

top_logprobsnumber · max: 20 · nullableOptional

Other propertiesnumber · min: -100 · max: 100Optional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

min_pnumber · min: 0.001 · max: 0.999Optional

A number between 0.001 and 0.999 that can be used as an alternative to top_p and top_k.

top_knumberOptional

Only sample from the top K options for each subsequent token. Used to remove "long tail" low probability responses. Recommended for advanced use cases only. You usually only need to use temperature.

repetition_penaltynumber · nullableOptional

A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: meta-llama/Llama-3.3-70B-Instruct-Turbo

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

repetition_penaltynumber · nullableOptional

A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.

logprobsboolean · nullableOptional

Whether to return log probabilities of the output tokens or not. If True, returns the log probabilities of each output token returned in the content of message.

top_logprobsnumber · max: 20 · nullableOptional

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: alibaba/qwen3-max-instruct

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

repetition_penaltynumber · nullableOptional

A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.

logprobsboolean · nullableOptional

Whether to return log probabilities of the output tokens or not. If True, returns the log probabilities of each output token returned in the content of message.

top_logprobsnumber · max: 20 · nullableOptional

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: alibaba/qwen3.5-plus-20260218

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

logprobsboolean · nullableOptional

Whether to return log probabilities of the output tokens or not. If True, returns the log probabilities of each output token returned in the content of message.

top_logprobsnumber · max: 20 · nullableOptional

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: qwen-turbo

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

repetition_penaltynumber · nullableOptional

A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.

logprobsboolean · nullableOptional

Whether to return log probabilities of the output tokens or not. If True, returns the log probabilities of each output token returned in the content of message.

top_logprobsnumber · max: 20 · nullableOptional

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: alibaba/qwen3-max-preview

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequired

The type of the content part.

Possible values:

file_datastringOptional

filenamestringOptional

The file name specified by the user. This name can be used to reference the file when interacting with the model, especially if multiple files are uploaded.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

contentany ofRequired

The contents of the developer message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

rolestring · enumRequired

The role of the author of the message — in this case, the developer.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

echobooleanOptional

If True, the response will contain the prompt. Can be used with logprobs to return prompt logprobs.

min_pnumber · min: 0.001 · max: 0.999Optional

A number between 0.001 and 0.999 that can be used as an alternative to top_p and top_k.

top_knumberOptional

Only sample from the top K options for each subsequent token. Used to remove "long tail" low probability responses. Recommended for advanced use cases only. You usually only need to use temperature.

repetition_penaltynumber · nullableOptional

A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.

Other propertiesnumber · min: -100 · max: 100Optional

ninteger · min: 1 · nullableOptional

How many chat completion choices to generate for each input message. Note that you will be charged based on the number of generated tokens across all of the choices. Keep n as 1 to minimize costs.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: deepseek/deepseek-reasoner-v3.1-terminus

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

urlstring · uriRequired

Either a URL of the image or the base64 encoded image data.

detailstring · enumOptional

Specifies the detail level of the image. Currently supports JPG/JPEG, PNG, GIF, and WEBP formats.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

ninteger · min: 1 · nullableOptional

How many chat completion choices to generate for each input message. Note that you will be charged based on the number of generated tokens across all of the choices. Keep n as 1 to minimize costs.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: google/gemini-2.0-flash

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

urlstring · uriRequired

Either a URL of the image or the base64 encoded image data.

detailstring · enumOptional

Specifies the detail level of the image. Currently supports JPG/JPEG, PNG, GIF, and WEBP formats.

Possible values:

typestring · enumRequired

The type of the content part.

Possible values:

file_datastringOptional

filenamestringOptional

The file name specified by the user. This name can be used to reference the file when interacting with the model, especially if multiple files are uploaded.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: baidu/ernie-4.5-vl-424b-a47b

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequiredPossible values:

contentany ofRequired

stringOptional

itemsone ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

typestring · enumRequired

The type of the image.

Possible values:

media_typestring · enumRequired

The media type of the image.

Possible values:

datastringRequired

The base64 encoded image data.

typestring · enumRequiredPossible values:

thinkingstringRequired

signaturestringRequired

typestring · enumRequiredPossible values:

tool_use_idstringRequired

is_errorbooleanOptional

contentany ofOptional

stringOptional

itemsone ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

typestring · enumRequired

The type of the image.

Possible values:

media_typestring · enumRequired

The media type of the image.

Possible values:

datastringRequired

The base64 encoded image data.

idstringRequired

Other propertiesany · nullableOptional

namestringRequired

typestring · enumRequiredPossible values:

datastringRequired

typestring · enumRequiredPossible values:

sourceany ofRequired

typestring · enumRequiredPossible values:

media_typestring · enumRequiredPossible values:

datastringRequired

typestring · enumRequiredPossible values:

datastringRequired

Other propertiesstringOptional

stop_sequencesstring[]Optional

Custom text sequences that will cause the model to stop generating.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

systemstringOptional

A system prompt is a way of providing context and instructions to Claude, such as specifying a particular goal or role.

namestringRequired

Name of the tool.

descriptionstringOptional

typestring · enumRequiredPossible values:

propertiesany · nullableOptional

Other propertiesany · nullableOptional

typestring · enumRequiredPossible values:

namestringRequired

typestring · enumRequiredPossible values:

budget_tokensinteger · min: 1024Required

typestring · enumRequiredPossible values:

max_tokensnumberOptional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

Default: 32000

temperaturenumber · max: 1Optional

top_pnumber · max: 1Optional

top_knumberOptional

Only sample from the top K options for each subsequent token. Used to remove "long tail" low probability responses. Recommended for advanced use cases only. You usually only need to use temperature.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: anthropic/claude-sonnet-4

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

urlstring · uriRequired

Either a URL of the image or the base64 encoded image data.

detailstring · enumOptional

Specifies the detail level of the image. Currently supports JPG/JPEG, PNG, GIF, and WEBP formats.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

repetition_penaltynumber · nullableOptional

A number that controls the diversity of generated text by reducing the likelihood of repeated sequences. Higher values decrease repetition.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: alibaba/qwen3-vl-32b-thinking

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequiredPossible values:

contentany ofRequired

stringOptional

itemsone ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

typestring · enumRequired

The type of the image.

Possible values:

media_typestring · enumRequired

The media type of the image.

Possible values:

datastringRequired

The base64 encoded image data.

typestring · enumRequiredPossible values:

thinkingstringRequired

signaturestringRequired

typestring · enumRequiredPossible values:

tool_use_idstringRequired

is_errorbooleanOptional

contentany ofOptional

stringOptional

itemsone ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

typestring · enumRequired

The type of the image.

Possible values:

media_typestring · enumRequired

The media type of the image.

Possible values:

datastringRequired

The base64 encoded image data.

idstringRequired

Other propertiesany · nullableOptional

namestringRequired

typestring · enumRequiredPossible values:

datastringRequired

typestring · enumRequiredPossible values:

sourceany ofRequired

typestring · enumRequiredPossible values:

media_typestring · enumRequiredPossible values:

datastringRequired

typestring · enumRequiredPossible values:

datastringRequired

Other propertiesstringOptional

stop_sequencesstring[]Optional

Custom text sequences that will cause the model to stop generating.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

systemstringOptional

A system prompt is a way of providing context and instructions to Claude, such as specifying a particular goal or role.

namestringRequired

Name of the tool.

descriptionstringOptional

typestring · enumRequiredPossible values:

propertiesany · nullableOptional

Other propertiesany · nullableOptional

typestring · enumRequiredPossible values:

namestringRequired

typestring · enumRequiredPossible values:

budget_tokensinteger · min: 1024Required

typestring · enumRequiredPossible values:

max_tokensnumberOptional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

Default: 32000

temperaturenumber · max: 1Optional

top_pnumber · max: 1Optional

top_knumberOptional

Only sample from the top K options for each subsequent token. Used to remove "long tail" low probability responses. Recommended for advanced use cases only. You usually only need to use temperature.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: anthropic/claude-opus-4.1

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

urlstring · uriRequired

Either a URL of the image or the base64 encoded image data.

detailstring · enumOptional

Specifies the detail level of the image. Currently supports JPG/JPEG, PNG, GIF, and WEBP formats.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

ninteger · min: 1 · nullableOptional

How many chat completion choices to generate for each input message. Note that you will be charged based on the number of generated tokens across all of the choices. Keep n as 1 to minimize costs.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: google/gemini-2.5-pro

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

modelstring · enumRequiredPossible values:

rolestring · enumRequiredPossible values:

contentany ofRequired

stringOptional

itemsone ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

typestring · enumRequired

The type of the image.

Possible values:

media_typestring · enumRequired

The media type of the image.

Possible values:

datastringRequired

The base64 encoded image data.

typestring · enumRequiredPossible values:

thinkingstringRequired

signaturestringRequired

typestring · enumRequiredPossible values:

tool_use_idstringRequired

is_errorbooleanOptional

contentany ofOptional

stringOptional

itemsone ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

typestring · enumRequired

The type of the image.

Possible values:

media_typestring · enumRequired

The media type of the image.

Possible values:

datastringRequired

The base64 encoded image data.

idstringRequired

Other propertiesany · nullableOptional

namestringRequired

typestring · enumRequiredPossible values:

datastringRequired

typestring · enumRequiredPossible values:

sourceany ofRequired

typestring · enumRequiredPossible values:

media_typestring · enumRequiredPossible values:

datastringRequired

typestring · enumRequiredPossible values:

datastringRequired

Other propertiesstringOptional

stop_sequencesstring[]Optional

Custom text sequences that will cause the model to stop generating.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

systemstringOptional

A system prompt is a way of providing context and instructions to Claude, such as specifying a particular goal or role.

namestringRequired

Name of the tool.

descriptionstringOptional

typestring · enumRequiredPossible values:

propertiesany · nullableOptional

Other propertiesany · nullableOptional

typestring · enumRequiredPossible values:

namestringRequired

typestring · enumRequiredPossible values:

budget_tokensinteger · min: 1024Required

typestring · enumRequiredPossible values:

max_tokensnumberOptional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

Default: 32000

temperaturenumber · max: 1Optional

top_pnumber · max: 1Optional

top_knumberOptional

Only sample from the top K options for each subsequent token. Used to remove "long tail" low probability responses. Recommended for advanced use cases only. You usually only need to use temperature.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: anthropic/claude-haiku-4.5

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

Body

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

urlstring · uriRequired

Either a URL of the image or the base64 encoded image data.

detailstring · enumOptional

Specifies the detail level of the image. Currently supports JPG/JPEG, PNG, GIF, and WEBP formats.

Possible values:

typestring · enumRequired

The type of the content part.

Possible values:

dataany ofRequired

Either a URL of the audio or the base64 encoded audio data.

string · uriOptional

stringOptional

formatstring · enumRequired

The format of the encoded audio data. Currently supports "wav" and "mp3".

Possible values:

typestring · enumRequired

The type of the content part.

Possible values:

urlstring · uriRequired

Either a URL of the video or the base64 encoded video data.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

idstringRequired

Unique identifier for a previous audio response from the model.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

formatstring · enumRequired

Specifies the output audio format. Must be one of wav, mp3, flac, opus, or pcm16.

Possible values:

voiceany ofRequired

The voice the model uses to respond. Supported voices are alloy, ash, ballad, coral, echo, fable, nova, onyx, sage, and shimmer.

string · enumOptionalPossible values:

stringOptional

itemsstring · enumOptionalPossible values:

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

Other propertiesnumber · min: -100 · max: 100Optional

logprobsboolean · nullableOptional

Whether to return log probabilities of the output tokens or not. If True, returns the log probabilities of each output token returned in the content of message.

top_logprobsnumber · max: 20 · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

enable_thinkingbooleanOptional

Specifies whether to use the thinking mode.

Default: false

thinking_budgetinteger · min: 1Optional

The maximum reasoning length, effective only when enable_thinking is set to true.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: alibaba/qwen3.5-omni-plus

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

post

/v1/chat/completions

200Success

modelstring · enumRequiredPossible values:

rolestring · enumRequired

The role of the author of the message — in this case, the user

Possible values:

contentany ofRequired

The contents of the user message.

stringOptional

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

typestring · enumRequiredPossible values:

urlstring · uriRequired

Either a URL of the image or the base64 encoded image data.

detailstring · enumOptional

Specifies the detail level of the image. Currently supports JPG/JPEG, PNG, GIF, and WEBP formats.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the system.

Possible values:

contentany ofRequired

The contents of the system message.

stringOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the tool.

Possible values:

contentstringRequired

The contents of the tool message.

tool_call_idstringRequired

Tool call that this message is responding to.

namestring · nullableOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

rolestring · enumRequired

The role of the author of the message — in this case, the Assistant.

Possible values:

contentany ofOptional

The contents of the Assistant message. Required unless tool_calls or function_call is specified.

stringOptional

The contents of the Assistant message.

itemsany ofOptional

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

refusalstringRequired

The refusal message generated by the model.

typestring · enumRequired

The type of the content part.

Possible values:

namestringOptional

An optional name for the participant. Provides the model information to differentiate between participants of the same role.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

argumentsstringRequired

refusalstring · nullableOptional

The refusal message by the Assistant.

max_completion_tokensinteger · min: 1Optional

An upper bound for the number of tokens that can be generated for a completion, including visible output tokens and reasoning tokens.

max_tokensnumber · min: 1Optional

The maximum number of tokens that can be generated in the chat completion. This value can be used to control costs for text generated via API.

streambooleanOptional

If set to True, the model response data will be streamed to the client as it is generated using server-sent events.

Default: false

include_usagebooleanRequired

ninteger · min: 1 · nullableOptional

How many chat completion choices to generate for each input message. Note that you will be charged based on the number of generated tokens across all of the choices. Keep n as 1 to minimize costs.

temperaturenumber · max: 2Optional

top_pnumber · min: 0.01 · max: 1Optional

stopany ofOptional

Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.

stringOptional

string[]Optional

any · nullableOptional

frequency_penaltynumber · min: -2 · max: 2 · nullableOptional

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.

typestring · enumRequired

The type of the predicted content you want to provide.

Possible values:

contentany ofRequired

The content that should be matched when generating a model response. If generated tokens would match this content, the entire model response can be returned much more quickly.

stringOptional

The content used for a Predicted Output. This is often the text of a file you are regenerating with minor changes.

typestring · enumRequired

The type of the content part.

Possible values:

textstringRequired

The text content.

presence_penaltynumber · min: -2 · max: 2 · nullableOptional

Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.

seedinteger · min: 1Optional

This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

response_formatone ofOptional

An object specifying the format that the model must output.

typestring · enumRequired

The type of response format being defined. Always text.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_object.

Possible values:

typestring · enumRequired

The type of response format being defined. Always json_schema.

Possible values:

namestringRequired

The name of the response format. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

strictboolean · nullableOptional

descriptionstringOptional

A description of what the response format is for, used by the model to determine how to respond in the format.

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

descriptionstringOptional

A description of what the function does, used by the model to choose when and how to call the function.

namestringRequired

The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.

Other propertiesany · nullableOptional

The parameters the functions accepts, described as a JSON Schema object.

strictboolean · nullableOptional

tool_choiceany ofOptional

string · enumOptional

Possible values:

typestring · enumRequired

The type of the tool. Currently, only function is supported.

Possible values:

namestringRequired

The name of the function to call.

parallel_tool_callsbooleanOptional

Whether to enable parallel function calling during tool use.

Responses

200Success

idstringRequired

A unique identifier for the chat completion.

Example: chatcmpl-CQ9FPg3osank0dx0k46Z53LTqtXMl

objectstring · enumRequired

The object type.

Example: chat.completionPossible values:

creatednumberRequired

The Unix timestamp (in seconds) of when the chat completion was created.

Example: 1762343744

indexnumberRequired

The index of the choice in the list of choices.

Example: 0

rolestringRequired

The role of the author of this message.

Example: assistant

contentstringRequired

The contents of the message.

Example: Hello! I'm just a program, so I don't have feelings, but I'm here and ready to help you. How can I assist you today?

refusalstring · nullableOptional

The refusal message generated by the model.

typestring · enumRequired

The type of the URL citation. Always url_citation.

Possible values:

end_indexintegerRequired

The index of the last character of the URL citation in the message.

start_indexintegerRequired

The index of the first character of the URL citation in the message.

titlestringRequired

The title of the web resource.

urlstringRequired

The URL of the web resource.

idstringRequired

Unique identifier for this audio response.

datastringRequired

Base64 encoded audio bytes generated by the model, in the format specified in the request.

transcriptstringRequired

Transcript of the audio generated by the model.

expires_atintegerRequired

The Unix timestamp (in seconds) for when this audio response will no longer be accessible on the server for use in multi-turn conversations.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

argumentsstringRequired

namestringRequired

The name of the function to call.

idstringRequired

The ID of the tool call.

typestring · enumRequired

The type of the tool.

Possible values:

inputstringRequired

The input for the custom tool call generated by the model.

namestringRequired

The name of the custom tool to call.

finish_reasonstring · enumRequired

Possible values:

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[]Required

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

bytesinteger[] · nullableOptional

logprobnumberRequired

The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.

tokenstringRequired

The token.

modelstringRequired

The model used for the chat completion.

Example: google/gemini-3-flash-preview

prompt_tokensnumberRequired

Number of tokens in the prompt.

Example: 137

completion_tokensnumberRequired

Number of tokens in the generated completion.

Example: 914

total_tokensnumberRequired

Total number of tokens used in the request (prompt + completion).

Example: 1051

accepted_prediction_tokensinteger · nullableOptional

When using Predicted Outputs, the number of tokens in the prediction that appeared in the completion.

audio_tokensinteger · nullableOptional

Audio input tokens generated by the model.

reasoning_tokensinteger · nullableOptional

Tokens generated by the model for reasoning.

rejected_prediction_tokensinteger · nullableOptional

audio_tokensinteger · nullableOptional

Audio input tokens present in the prompt.

cached_tokensinteger · nullableOptional

Cached tokens present in the prompt.

credits_usednumberRequired

The number of tokens consumed during generation.

Example: 120000

usd_spentnumberRequired

The total amount of money spent by the user in USD.

Example: 0.06

AI/ML API Documentation

Quickstart

API REFERENCES

Service Endpoints

All Model IDs

Full List of Model IDs

Text Models (LLM)

Image Models

Video Models

Voice/Speech Models

Speech-to-Text

Text-to-Speech

Voice Chat

Music Models

Vision Models

Optical Character Recognition (OCR)

3D-Generating Models

Embedding Models

Deprecated / No Longer Supported Models

Text Models (LLM)

qwen3-vl-32b-instruct

Model Overview

ernie-4.5-300b-a47b

Model Overview

ByteDance

gemini-3-1-flash-lite-preview

Model Overview

API Schema

Code Example

gemini-3-1-flash-lite-preview

Model Overview

API Schema

Code Example

All Model IDs

Full List of Model IDs

Text Models (LLM)

Image Models

Video Models

Voice/Speech Models

Speech-to-Text

Text-to-Speech

Voice Chat

Music Models

Vision Models

Optical Character Recognition (OCR)

3D-Generating Models

Embedding Models

Deprecated / No Longer Supported Models

Text Models (LLM)

ernie-4.5-300b-a47b

Model Overview

qwen3-vl-32b-instruct

Model Overview

ByteDance

Service Endpoints

API Schema

Code Example

API Schema

Code Example

ernie-4.5-turbo-vl-32k

Model Overview

API Schema

Code Example

ernie-4.5-300b-a47b-paddle

Model Overview

API Schema

Code Example

Gryphe

Cohere

Documentation Map

Browse Models

Going Deeper

Have a Minute? Help Make the Docs Better!

DeepSeek

Google

Alibaba Cloud

Anthracite

Baidu

Meta

MiniMax