Ctrlk
API KeyModelsPlaygroundGitHubGet Support

AutoGPT

About

AutoGPT is an open-source platform designed to help you build, test, and run AI agents using a no-code visual interface. It allows users to link LLMs with tools, memory, planning modules, and action chains. By configuring block-based workflows, you can create custom agents that reason, plan, and act in multi-step environments.

In this guide, you'll learn how to connect AutoGPT with high-performance language models from AI/ML API for use in AI-driven text generation tasks.

Prerequisites

Before proceeding, ensure:

Step-by-Step Setup

🥇 Step 1. Install and Launch AutoGPT Locally

Use the latest official guide published on the AutoGPT documentation site: AutoGPT Getting Started Guide

💡 Tip: Always refer to the most recent version of the guide to avoid setup issues or deprecated steps.

Make sure you're running AutoGPT on http://localhost:3000.

🥈 Step 2. Open the Visual Block Builder

Before proceeding, make sure you're logged in to your AI/ML API account or create an account if you haven't already: aimlapi.com

Once AutoGPT is running:

🔗 Open: http://localhost:3000/build or click "Build" from the navigation bar.

💡 Tip: This is your no-code playground to configure agents and workflows.

🥉 Step 3. Click “Blocks” on the Left Sidebar

  • Find the left panel.

  • Click the button labeled "Blocks".

This shows you all available functional blocks (including LLMs, tools, memory, etc.)

🔍 Step 4. Search for “AI Text Generator”

In the search bar:

  • Type: ai text generator

  • Click on the AI Text Generator block when it appears.

🧠 This block lets you plug in a language model for text completions, prompts, and chat flows.

🤖 Step 5. Select One of the Currently Supported AIMLAPI Models

The screenshot may show more model options.

Some models shown in the screenshot are outdated and are no longer supported. Because models evolve and become outdated quickly, the only currently supported model is meta-llama/Llama-3.3-70B-Instruct-Turbo.

We plan to refresh the available model list soon.

Click the block to configure it, then select one of the currently supported AIMLAPI models in the model selection field.

🔑 Step 6. Enter Your Prompt and API Credentials

In the AI Text Generator block:

  1. Set Prompt: Type any message you want the model to respond to.

  2. Set API Key:

    • Enter your AI/ML API key.

💡 Get your API key here: aimlapi.com/app/keys

🎉 Step 7. Done – You’re All Set!

Now that you’ve configured the prompt, selected a model, and added your API key — let’s finalize and run your agent in AutoGPT.

✅ 1. Save your Agent

Before running, make sure to save your current block configuration as an agent:

  1. Click the "Save" button at the top-right of the builder interface.

  2. In the popup, enter a name for your agent (e.g., aimlapi_test_agent).

  3. Click "Save Agent" to confirm.

💡 Saving your agent allows you to reuse it, schedule runs, or chain it into larger workflows with memory, tools, and action blocks.

▶️ 2. Run your Agent

After saving, you can now launch the agent:

  1. Press the "Run" button next to your agent on the workspace screen.

  2. AutoGPT will trigger the AI Text Generator block and initiate a request to the AI/ML API model.

⏱️ At this point, the system will send your prompt to the selected model and wait for a response.

🧾 3. View the Output

  1. Navigate to the "Output" Panel. At your AI/ML API block, locate the "Output" panel below.2. You'll see the response returned by the AI/ML API model.

  2. You can copy the result, export it, or pass it into further blocks (like analysis, memory, or a webhook).

🎉 That’s it! Your AutoGPT agent is now generating text using AI/ML API’s powerful language models.

💡 You can now expand your agent by chaining the AI Text Generator block to:

  • 🔧 Tools – call external APIs, perform web scraping, manage files.

  • 🧠 Memory – store and reuse past interactions for contextual reasoning.

  • ⚙️ Actions / Chains – create complex behavior flows and intelligent pipelines.

More

For further information about the framework, please check the official AutoGPT documentation.

PreviousAiderNextBrowser Use

Last updated

Was this helpful?