## Create task
`POST /v2/video_generation`
```bash cURL theme={null}
curl --request POST \
--url https://www.anyfast.ai/v2/video_generation \
--header 'Authorization: Bearer YOUR_API_KEY' \
--header 'Content-Type: application/json' \
--data '{
"model": "MiniMax-H3",
"content": [
{
"type": "text",
"text": "史诗级太空歌剧院线预告：女舰长独自站在巨大观景窗前，最后一支舰队正在集结并跃迁离去。"
}
],
"resolution": "2K",
"duration": 5,
"ratio": "16:9"
}'
```
```json 200 theme={null}
{
"task_id": "424010985738629"
}
```
## Request headers
Bearer authentication in the form `Bearer YOUR_API_KEY`.
Request body format.
## Request parameters
Use `MiniMax-H3`.
Multimodal content parts. Each item uses `text`, `image_url`, `video_url`, or `audio_url`.
Content type: `text`, `image_url`, `video_url`, or `audio_url`.
Text prompt. A text part is required for video generation.
Media role such as `first_frame`, `last_frame`, `reference_image`, `reference_video`, or `reference_audio`.
Public image URL. Base64 is not supported.
Public video URL for multimodal reference input.
Public audio URL for multimodal reference input.
Currently `2K`.
Video duration in seconds, from 4 to 15.
`adaptive`, `21:9`, `16:9`, `4:3`, `1:1`, `3:4`, or `9:16`. Text-to-video requires an explicit ratio.
Optional callback URL for task status updates.
Whether to add an AIGC watermark.
### Content rules
* Text-to-video accepts one text part.
* Image-to-video accepts a first frame, last frame, or both. Use `first_frame` and `last_frame` roles when supplying both.
* Reference-to-video can combine reference images, video, and audio, but must include at least one reference image or video; audio alone is not sufficient.
* First/last-frame roles cannot be mixed with reference-media roles.
* The total request body is limited to 64 MB. Images support JPG, JPEG, PNG, WEBP, HEIC, and HEIF; each image is limited to 30 MB.
* Reference videos support MP4/MOV, H.264/H.265, and AAC/MP3 audio; each is limited to 50 MB and 2-15 seconds. Reference audio supports WAV/MP3 and is limited to 15 MB and 2-15 seconds.
## Response
Video generation task ID. Use it with the query endpoint.
## Query task
`GET /v2/query/video_generation/{task_id}`
```bash cURL theme={null}
curl --request GET \
--url https://www.anyfast.ai/v2/query/video_generation/TASK_ID \
--header 'Authorization: Bearer YOUR_API_KEY'
```
```json 200 theme={null}
{
"task": {
"id": "424010985738629",
"model": "MiniMax-H3",
"status": "succeeded",
"created_at": 1785125529,
"updated_at": 1785125946,
"content": {
"url": "https://cdn.hailuoai.com/example/video.mp4"
},
"resolution": "2K",
"duration": 5,
"usage": {
"total_seconds": 5,
"input_seconds": 0,
"output_seconds": 5,
"image_count": 0
},
"ratio": "16:9",
"task_type": "generation"
}
}
```
## Request headers
Bearer authentication in the form `Bearer YOUR_API_KEY`.
## Path parameters
Video generation task ID returned by the create endpoint.
## Response fields
Queried video generation task.
Task ID.
Model used for the task.
Task status: `queued`, `running`, `succeeded`, `failed`, or `expired`.
Task creation time as a Unix timestamp.
Task update time as a Unix timestamp.
Generated video URL when the task succeeds.
Output resolution.
Output duration in seconds.
Usage metrics for the task.
Total billed or processed seconds.
Input media duration in seconds.
Output video duration in seconds.
Number of input images.
Output aspect ratio.
Task type, such as `generation`.