Volcengine Create Video Generation Task

POST /api/v3/contents/generations/tasks

Video generation endpoint in Volcengine Ark native format, connected to the upstream Volcengine Ark Seedance model family. The model asynchronously generates a video from the provided text, image, audio, and video inputs. This endpoint returns a task ID after submission; use Query Video Generation Task to poll the task status and retrieve the generated video URL.

⚠️ This is an asynchronous endpoint. A successful submission only means the task has been created. Poll the task status using the returned task ID.

Request Parameters

The request body follows the Volcengine Ark native format and passes through all top-level parameters. Supported parameters and values vary by model.

Name	Type	Required	Description
model	string	Required	Model ID or Endpoint ID
content	array<object>	Required	Input content list, supporting text, image, audio, video, and draft-task elements
content[].type	string	Required	Content type: text, image_url, video_url, audio_url, draft_task
content[].text	string	Optional	Text prompt (when type=text) describing the desired video
content[].image_url	object	Optional	Image object (when type=image_url)
content[].image_url.url	string	Optional	Image source: image URL, Base64 (data:image/<format>;base64,<...>), or asset ID (asset://<ASSET_ID>)
content[].video_url	object	Optional	Video object (when type=video_url)
content[].video_url.url	string	Optional	Video source: video URL or asset ID (asset://<ASSET_ID>)
content[].audio_url	object	Optional	Audio object (when type=audio_url)
content[].audio_url.url	string	Optional	Audio source: audio URL, Base64 (data:audio/<format>;base64,<...>), or asset ID
content[].draft_task	object	Optional	Draft-task object (when type=draft_task)
content[].draft_task.id	string	Optional	Draft task ID; generates a final video based on the draft
content[].role	string	Optional	Role/purpose: first_frame, last_frame, reference_image, reference_video, reference_audio
omni_reference_task_type	string	Optional	Omni reference task type hint, default auto; options auto, reference, edit, extend (Seedance 2.5 only)
resolution	string	Optional	Video resolution: 480p, 720p, 1080p, 4k (values vary by model)
ratio	string	Optional	Aspect ratio: 16:9, 4:3, 1:1, 3:4, 9:16, 21:9, adaptive
duration	integer	Optional	Video duration (seconds); -1 lets the model choose; mutually exclusive with frames
frames	integer	Optional	Frame count; integers in [29, 289] matching 25 + 4n; takes priority over duration
generate_audio	boolean	Optional	Whether to generate audio, default true
watermark	boolean	Optional	Whether to include the AI generated watermark, default false
output_format	string	Optional	Output format: mp4 (default), mov (Seedance 2.5 only)
seed	integer	Optional	Random seed, default -1 (random); range [-1, 2147483647]
camera_fixed	boolean	Optional	Whether to fix the camera, default false
return_last_frame	boolean	Optional	Whether to return the last frame of the generated video, default false
draft	boolean	Optional	Whether to enable draft mode to generate a preview video, default false (Seedance 1.5 pro only)
service_tier	string	Optional	Service tier: default (online, default), flex (offline)
callback_url	string	Optional	Callback URL notified when the task status changes
execution_expires_after	integer	Optional	Task timeout threshold (seconds), default 172800 (48 hours); range [3600, 259200]
priority	integer	Optional	Execution priority, default 0; range [0, 9], higher is prioritized (Seedance 2.5 / 2.0 series only)
safety_identifier	string	Optional	Unique end-user identifier, an ASCII string up to 64 characters
tools	array<object>	Optional	Tool configuration (Seedance 2.5 / 2.0 series only)
tools[].type	string	Optional	Tool type, e.g. web_search (web search)

Request Example

{
  "body": {
    "model": "doubao-seedance-1-5-pro-251215",
    "content": [
      {
        "type": "text",
        "text": "A kitten yawning at the camera"
      }
    ],
    "resolution": "720p",
    "ratio": "16:9",
    "duration": 5,
    "seed": 11,
    "camera_fixed": false,
    "watermark": true
  }
}

Response Fields (200)

Name	Type	Description
id	string	Video generation task ID
task_id	string	Video generation task ID (same as id)
object	string	Object type, always video
model	string	Model name
status	string	Task status
created_at	integer	Task creation timestamp (seconds)

Response Example

{
  "id": "cgt-2026xxxx-xxxxx",
  "task_id": "cgt-2026xxxx-xxxxx",
  "object": "video",
  "model": "doubao-seedance-1-5-pro-251215",
  "status": "queued",
  "created_at": 1757222400
}