---
title: "Avatar Video"
method: POST
path: "/avatars/video"
tags: ["Avatar"]
---

# Avatar Video

`POST /avatars/video`

## Request

- Base URL: `https://genai-avatars-api.picsart.io/v1`
- URL: `https://genai-avatars-api.picsart.io/v1/avatars/video`
- Auth: API key in header `X-Picsart-API-Key`

Animates a person from a single image so that they lip-sync a provided audio track. The process is asynchronous and will require an additional call to get the result. See below for more details.

## Request body

- object
  - `image_url` string, uri, required: Source image URL of the person to animate.
  - `audio_url` string, uri, required: URL of the audio track the avatar lip-syncs to. The length of the track drives the duration of the generated video, and with it the price. Depending on the selected model, WAV and MP3 may be the only accepted formats, and the track may be capped at 60 seconds for 720p or 30 seconds for 1080p. The boreal model accepts at most 20 seconds of audio; h3-max-lip-sync requires a track between 5 and 14.8 seconds.
  - `prompt` string: Text prompt guiding motion, emotion and camera movement. Required by the boreal model, which accepts at most 5000 characters; ignored by the fabric and h3-max-lip-sync models.
  - `quality` '480p' | '720p' | '1080p' | '4k': Video output quality/resolution. Mapped to the closest tier the underlying model supports. Options are as follows: * 480p * 720p * 1080p * 4k
  - `model` 'urn:air:heygen:model:heygen:avatar-iv@1' | 'urn:air:kling:model:kling:kling-avatar-std@1' | 'urn:air:kling:model:kling:kling-avatar-pro@1' | 'urn:air:bytedance:model:bytedance:omnihuman-1.5@1' | 'urn:air:creatify:model:creatify:aurora@1' | 'urn:air:creatify:model:creatify:boreal@1' | 'urn:air:veed:model:veed:fabric-1.0@1' | 'urn:air:veed:model:veed:fabric-1.0-fast@1' | 'urn:air:minimax:model:minimax:h3-max-lip-sync@1', nullable: Optionally choose a specific AI model to use for this video generation. If not specified, the model configured for your API key is used, and where none is configured a default model will be applied. Please note that the default model may change over time as Picsart continues to improve performance and accuracy. Any change to the default will be made only after thorough testing and validation to ensure it delivers better results. If you require consistent behavior or wish to evaluate different models on your own, we recommend explicitly setting this parameter.

## Response `202`

Accepted.

- object
  - `status` 'processing' | 'error' | 'success': The status of the submitted request. Pay attention to handle these properly: * processing * error * success
  - `inference_id` string: The unique identifier of this inference. Use it to poll for the result.

## Other responses

- `400`: Bad Request
- `401`: Unauthorized
- `403`: Forbidden
- `404`: Not Found
- `405`: Method Not Allowed
- `413`: Request Entity Too Large
- `415`: Unsupported Media Type
- `429`: Too Many Requests
- `431`: Request Header Fields Too Large
- `500`: Internal Server Error
- `503`: Service Unavailable
- `529`: Overloaded. The model provider is overloaded and could not complete this request. Please retry shortly, or switch to a different model.

---

[API](https://skmtc.dev/picsart/apis/picsart-genai-avatars-api.md) - [All operations](https://skmtc.dev/picsart/apis/picsart-genai-avatars-api/llms.txt) - [OpenAPI document](https://skmtc.dev/picsart/apis/picsart-genai-avatars-api/revisions/42b369953263?raw)