> For the complete documentation index, see [llms.txt](https://docs.aimlapi.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.aimlapi.com/api-references/video-models/minimax-ai/recast.md).

# MiniMax H3 Max Recast

Model id: minimax/h3-max-recast. MiniMax H3 Max Recast replaces the main people in a 5-30 second video with the people from 1-4 reference photos, keeping the original motion, camera, cuts and sound, a

{% columns %}
{% column width="66.66666666666666%" %}
{% hint style="info" %}
This documentation is valid for the following list of our models:

* `minimax/h3-max-recast`
  {% endhint %}
  {% endcolumn %}

{% column width="33.33333333333334%" %} <a href="https://aimlapi.com/app/minimax/h3-max-recast" class="button primary">Try in Playground</a>
{% endcolumn %}
{% endcolumns %}

## Model Overview

MiniMax H3 Max Recast replaces the main people in a 5-30 second video with the people from 1-4 reference photos, keeping the original motion, camera, cuts and sound, at 768p or 1080p.

## Setup your API Key

If you don't have an API key yet, use our [Quickstart guide](https://docs.aimlapi.com/quickstart/setting-up).

## How to Make a Call

<details>

<summary>Step-by-Step Instructions</summary>

Generating a video involves sequentially calling two endpoints:

* Create and submit a video generation task. The response contains a generation ID.
* Poll the retrieval endpoint with that generation ID until the status is `completed`.

</details>

## API Schemas

### Create a video generation task and send it to the server

## POST /v2/video/generations

>

```json
{"openapi":"3.0.0","info":{"title":"AIML API","version":"1.0.0"},"servers":[{"url":"https://api.aimlapi.com"}],"paths":{"/v2/video/generations":{"post":{"operationId":"_v2_video_generations","requestBody":{"required":true,"content":{"application/json":{"schema":{"type":"object","properties":{"model":{"type":"string","enum":["minimax/h3-max-recast"]},"video_url":{"type":"string","format":"uri","description":"The source video, 5 to 30 seconds long, with no single shot longer than 15 seconds. The result keeps its length, motion, camera, cuts and sound."},"reference_image_urls":{"type":"array","items":{"type":"string","format":"uri"},"minItems":1,"maxItems":4,"description":"One photo per new person, 1 to 4 photos. By default they replace the main people in the video from left to right, in every shot they appear in."},"prompt":{"type":"string","maxLength":2000,"description":"Who becomes whom, or what else to keep or change."},"resolution":{"type":"string","enum":["768p","1080p"],"default":"1080p","description":"The resolution of the output video, where the number refers to the short side in pixels."},"seed":{"type":"integer","minimum":0,"description":"Varying the seed integer is a way to get different results for the same other request parameters. Using the same value for an identical request will produce similar results. If unspecified, a random number is chosen."},"enhance_realism":{"type":"boolean","default":false,"description":"More natural skin textures and better blending of the new people into the scene."},"webhook_url":{"type":"string","maxLength":2048,"format":"uri","description":"A public https address on port 443 or 8443 that gets a signed POST when the job ends. Needs a signing secret: POST /v1/webhooks/secret. On /v1/responses it needs `background: true` and does not work with `store: false`. On /v1/images/generations it runs the generation as a task on the models that support it (the others answer 400 webhook_url_not_supported): the answer is { id, status } and the result is read with GET /v1/images/generations/{id}."}},"required":["model","video_url","reference_image_urls"],"title":"minimax/h3-max-recast"}}}},"responses":{"200":{"content":{"application/json":{"schema":{"type":"object","properties":{"id":{"type":"string","description":"The ID of the generated video."},"status":{"type":"string","enum":["queued","generating","completed","error"],"description":"The current status of the generation task."},"video":{"anyOf":[{"type":"object","properties":{"url":{"type":"string","format":"uri","description":"The URL where the file can be downloaded from."}},"required":["url"]},{"type":"array","items":{"type":"object","properties":{"url":{"type":"string","format":"uri","description":"The URL where the file can be downloaded from."}},"required":["url"]}},{"nullable":true}]},"error":{"type":"object","nullable":true,"properties":{"name":{"type":"string"},"message":{"type":"string"}},"required":["name","message"],"description":"Description of the error, if any."},"meta":{"type":"object","nullable":true,"properties":{"usage":{"type":"object","nullable":true,"properties":{"credits_used":{"type":"number","description":"The number of tokens consumed during generation."},"usd_spent":{"type":"number","description":"The total amount of money spent by the user in USD."}},"required":["credits_used","usd_spent"]}},"description":"Additional details about the generation."}},"required":["id","status"]}}},"description":"Successful response."}}}}}}
```

### Retrieve the generated video from the server

Poll this endpoint with the `generation_id` returned by the submit request. When the status is `completed`, the response includes the generated video URL.

## GET /v2/video/generations

>

```json
{"openapi":"3.0.0","info":{"title":"AIML API","version":"1.0.0"},"servers":[{"url":"https://api.aimlapi.com"}],"security":[{"access-token":[]}],"components":{"securitySchemes":{"access-token":{"scheme":"bearer","bearerFormat":"<YOUR_AIMLAPI_KEY>","type":"http","description":"Bearer key","in":"header"}}},"paths":{"/v2/video/generations":{"get":{"operationId":"_v2_video_generations","parameters":[{"name":"generation_id","required":true,"in":"query","schema":{"type":"string"}}],"responses":{"200":{"content":{"application/json":{"schema":{"type":"object","properties":{"id":{"type":"string","description":"The ID of the generated video."},"status":{"type":"string","enum":["queued","generating","completed","error"],"description":"The current status of the generation task."},"video":{"type":"object","nullable":true,"properties":{"url":{"type":"string","format":"uri","description":"The URL where the file can be downloaded from."}},"required":["url"]},"error":{"type":"object","nullable":true,"properties":{"name":{"type":"string"},"message":{"type":"string"}},"required":["name","message"],"description":"Description of the error, if any."},"meta":{"type":"object","nullable":true,"properties":{"usage":{"type":"object","nullable":true,"properties":{"credits_used":{"type":"number","description":"The number of tokens consumed during generation."}},"required":["credits_used"]}},"description":"Additional details about the generation."}},"required":["id","status"]}}}}}}}}}
```

## Code Example

The examples submit a generation task and poll every **15 seconds** until the final result is available.

{% tabs %}
{% tab title="Python" %}
{% code overflow="wrap" %}

```python
import requests
import time

api_key = "<YOUR_AIMLAPI_KEY>"
url = "https://api.aimlapi.com/v2/video/generations"
headers = {
    "Authorization": f"Bearer {api_key}",
    "Content-Type": "application/json",
}

generation = requests.post(url, headers=headers, json={'model': 'minimax/h3-max-recast', 'video_url': 'https://example.com/input.jpg', 'reference_image_urls': ['https://example.com/input.jpg'], 'prompt': 'Describe what you want the model to generate.'})
generation.raise_for_status()
generation_id = generation.json()["id"]

while True:
    result = requests.get(
        url,
        headers=headers,
        params={"generation_id": generation_id},
    )
    result.raise_for_status()
    data = result.json()
    if data.get("status") not in ("queued", "generating"):
        print(data)
        break
    time.sleep(15)
```

{% endcode %}
{% endtab %}

{% tab title="JavaScript" %}
{% code overflow="wrap" %}

```javascript
const apiKey = "<YOUR_AIMLAPI_KEY>";
const url = "https://api.aimlapi.com/v2/video/generations";
const headers = {
  "Authorization": `Bearer ${apiKey}`,
  "Content-Type": "application/json",
};

const generation = await fetch(url, {
  method: "POST",
  headers,
  body: JSON.stringify({
  "model": "minimax/h3-max-recast",
  "video_url": "https://example.com/input.jpg",
  "reference_image_urls": [
    "https://example.com/input.jpg"
  ],
  "prompt": "Describe what you want the model to generate."
}),
});
const { id: generationId } = await generation.json();

while (true) {
  const response = await fetch(
    `${url}?generation_id=${encodeURIComponent(generationId)}`,
    { headers },
  );
  const data = await response.json();
  if (!["queued", "generating"].includes(data.status)) {
    console.log(data);
    break;
  }
  await new Promise((resolve) => setTimeout(resolve, 15000));
}
```

{% endcode %}
{% endtab %}
{% endtabs %}

<details>

<summary>Response</summary>

{% code overflow="wrap" %}

```json
{
  "id": "60ac7c34-3224-4b14-8e7d-0aa0db708325",
  "status": "completed",
  "video": {
    "url": "https://cdn.aimlapi.com/generations/hedgehog/1759866285599-0cdfb138-c03a-49d4-a601-4f6413e27b15.mp4"
  },
  "error": {
    "name": "<name>",
    "message": "<message>"
  },
  "meta": {
    "usage": {
      "credits_used": 120000,
      "usd_spent": 0.06
    }
  },
  "model": "minimax/h3-max-recast"
}
```

{% endcode %}

</details>


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://docs.aimlapi.com/api-references/video-models/minimax-ai/recast.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.
