4Z API

tusheng video

Endpoint: POST https://4zapi.com/ent/v2/img2video

Generate a video from the first frame of the image.

Request body (application/json)

  • images (array, required)
  • model (string, required): Model name Values: viduq4-preview, viduq3-pro, viduq3-turbo, viduq2-pro, viduq2-turbo - viduq4-preview: Q4 preview; 3–16 s, resolution 540p / 720p / 1080p / 2K / 4K; audio is generated by default when audio is omitted - viduq3-turbo: faster than viduq3-pro - viduq3-pro: high-quality audio-video generation with more vivid results - viduq2-pro: Q2 pro, better visual quality - viduq2-turbo: Q2 turbo, faster generation Note: plain viduq2 is not supported for image-to-video and returns 400; use viduq2-pro or viduq2-turbo
  • prompt (string, required): Text prompt describing the video (optional). Max 5000 characters for viduq4-preview, 2000 for other models.
  • audio (string)
  • callback_url (string): Callback protocol: set callback_url when creating the task (POST). On status changes, Vidu sends the latest task status (same shape as the query task API). status: processing, success, failed. See callback signature docs for auth. Note: viduq4-preview does not support callbacks; poll the query task API instead.
  • duration (number)
  • is_rec (string)
  • meta_data (string): Metadata identifier as a JSON string (pass-through). You may use a custom format or the sample shape with Label, ContentProducer, etc. If empty, Vidu-generated metadata is used. Note: not supported by viduq4-preview.
  • movement_amplitude (string)
  • off_peak (string)
  • payload (string): Pass-through parameter. Not processed; forwarded as-is. Max 1048576 characters.
  • resolution (string)
  • seed (string): Random seed. If omitted or 0, a random value is used; otherwise the provided seed is used.
  • voice_id (string)
  • watermark (string): Whether to add a watermark. true: add watermark; false: no watermark. Note: watermark content is fixed ("AI generated") and off by default. Use watermarked_url from the query task API for watermarked video.
  • wm_position (string): Watermark position on the image: 1 top-left, 2 top-right, 3 bottom-right, 4 bottom-left. Default: 3.
  • wm_url (string): Watermark content as an image URL. If omitted, the default watermark "AI generated" is used.

Request example

{
  "audio": false,
  "duration": 5,
  "images": [
    "https://imageproxy.zhongzhuan.chat/api/proxy/image/6fb238cc01649cb7580fbb2dd58f3d6d.png"
  ],
  "model": "viduq2-pro",
  "movement_amplitude": "auto",
  "off_peak": false,
  "prompt": "镜头缓慢推进，展现画面中的细节，背景逐渐虚化",
  "resolution": "1080p",
  "seed": 12345,
  "watermark": false
}