Jimeng Video Generation
Use `POST /v1/videos` to call the Jimeng model and submit asynchronous video tasks.
https://zx1.deepwl.net/v1/videoscurl -X POST https://zx1.deepwl.net/v1/videos \
-H "Authorization: Bearer YOUR_API_KEY" \
-F "model=jimeng-video-3.0" \
-F "prompt=猫咪在草地上追逐蝴蝶" \
-F "seconds=5" \
-F "size=1280x720"{
"id": "video_123",
"task_id": "video_123",
"object": "video",
"model": "jimeng-video-3.0",
"status": "queued",
"progress": 0,
"created_at": 1712697600
}OpenAI-format entry point. If you need to use the unified video POST /v1/video/create, see Generate Video.
Jimeng currently uses POST /v1/videos to submit tasks, primarily via multipart/form-data.
- The reference image is passed as a single file through the
input_referencefield. - After a successful submission, the task
idandstatusare returned, and you can later poll the result via Task Status Query.
Supported Models
jimeng-video-3.0jimeng-video-2.0
Request Headers
Authorizationstring必填Authentication header. Uses a Bearer token, e.g. Bearer YOUR_API_KEY.
Request Body
Submitted as a multipart/form-data form; the reference image is uploaded as an image file directly.
modelstring必填Model name. Supports jimeng-video-3.0 and jimeng-video-2.0.
promptstring必填Prompt describing the video content. Supports Chinese and English.
input_referencefileReference image. Upload a single image file for image-to-video scenarios. Supported image formats include png, jpeg, jpg, and webp.
secondsstringVideo duration in seconds. Available values: 5, 10, 15.
sizestringVideo size. Available values: 720x1280 (portrait), 1280x720 (landscape). Defaults to 720x1280.
Request Examples
Response Examples
Current Rules
| Item | Rule |
|---|---|
| Default duration | If not explicitly provided, defaults to 5 seconds |
| Default size | If not explicitly provided, defaults to 720x1280 |
| Reference image | Upload a single image file via input_reference |
| Reference image formats | Supports png, jpeg, jpg, and webp |