多元探索 API多元探索 API
OpenAI Format

Vidu Video Generation

Use `POST /v1/videos` to call the Vidu model and submit asynchronous video tasks.

POSThttps://zx1.deepwl.net/v1/videos
Request
curl -X POST https://zx1.deepwl.net/v1/videos \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -F "model=viduq3-pro" \
  -F "prompt=A cat chasing butterflies on the grass" \
  -F "aspect_ratio=16:9" \
  -F "duration=5"
Response
{
  "id": "video_123",
  "object": "video",
  "model": "viduq3-pro",
  "status": "queued",
  "progress": 0,
  "created_at": 1712697600,
  "size": "1024x1808",
  "seconds": "5",
  "quality": "standard"
}

The Vidu channel video generation API, submitted as multipart/form-data, supports text-to-video and image-to-video; a single reference image file is uploaded through the images field. After a successful submission, the task id and status are returned; poll the result later with Task Status Query.

Supported Models

  • viduq3-pro: Efficiently generates high-quality audio and video content, making videos more vivid, more lifelike, and more three-dimensional
  • viduq2-pro: New model, good results, rich details
  • viduq2-turbo: New model, good results, fast generation

Request Headers

Authorizationstring必填

Authentication header. Uses a Bearer token, e.g. Bearer YOUR_API_KEY.

Request Body

modelstring必填

Model name. Supported values: viduq3-pro, viduq2-pro, viduq2-turbo.

promptstring必填

Text prompt. The text description for generating the video. If the is_rec recommended prompt parameter is used, the model will ignore the prompt entered in this parameter.

imagesfile

Reference image. Upload a single image file for image-to-video scenarios. Supports passing an image Base64 encoding or an image URL (ensure it is accessible). Only 1 image is supported. Supported image formats: png, jpeg, jpg, webp. The image aspect ratio must be less than 1:4 or greater than 4:1, and the image size must not exceed 50 MB.

aspect_ratiostring

Aspect ratio. Optional values include 16:9 (default), 9:16, 3:4, 4:3, 1:1. Note: 3:4 and 4:3 are only supported by q2 and q3 models.

audioboolean

Whether to use the audio-video direct output capability. false: no direct audio-video output, produces a silent video; true: direct audio-video output is required, producing a video with dialogue and background sound.

voice_idstring

Voice ID, not effective for q3 models. Used to determine the voice timbre in the video; when empty, the system will automatically recommend one.

is_recboolean

Whether to use recommended prompts. true: the system automatically recommends prompts and uses the prompt content to generate the video, recommended prompt count = 1; false: generate the video based on the input prompt.

bgmboolean

Whether to add background music to the generated video. true: the system will automatically select suitable music from the preset BGM library and add it; false: no BGM is added.

durationinteger

Video duration (seconds). The default for the viduq2 series is 5; optional: 1-10.

seedinteger

Random seed. When not passed by default or passed as 0, a random number will be used instead. If set manually, the configured seed will be used.

resolutionstring

Resolution. The default value depends on the model and video duration. viduq2-pro 1-10 seconds: default 720p, optional: 540p, 720p, 1080p; viduq2-turbo 1-10 seconds: default 720p, optional: 540p, 720p, 1080p.

payloadstring

Pass-through parameter. No processing is performed; data transmission only.

off_peakboolean

Off-peak mode. true: generate videos during off-peak hours; false: generate videos immediately.

watermarkboolean

Whether to add a watermark. true: add watermark; false: no watermark.

wm_positioninteger

Watermark position. 1: top-left; 2: top-right; 3: bottom-right; 4: bottom-left.

wm_urlstring

Watermark content, here it is an image URL. If not passed, the default watermark is used: content generated by AI.

meta_datastring

Metadata identifier, JSON-formatted string, pass-through field.

Request Examples

Response Examples

Response Fields

idstring

Task ID.

objectstring

Fixed as video.

modelstring

Model name.

statusstring

Task status. Optional values include queued (in queue), processing (processing), completed (completed), failed (failed), cancelled (cancelled).

progressinteger

Progress percentage (0-100).

created_atinteger

Creation time (Unix timestamp).

sizestring

Video size, for example 1024x1808.

secondsstring

Video duration (seconds).

qualitystring

Video quality, for example standard.