Skip to main content
POST
/
volc
/
v3
/
contents
/
generations
/
tasks
curl --request POST \ --url https://api.cometapi.com/volc/v3/contents/generations/tasks \ --header 'Authorization: Bearer <token>' \ --header 'Content-Type: application/json' \ --data ' { "model": "doubao-seedance-1-5-pro-251215", "content": [ { "type": "text", "text": "A paper boat drifts on a quiet lake at sunrise. --duration 5 --ratio 16:9" } ] } '
{
  "id": "021773297205999example"
}
Use this endpoint to start a ByteDance Seedance video task from text, or from text plus reference images. The request returns a task id immediately and the render continues asynchronously.

​
Choose the current model path

  • Use the Models page for current model availability and identifiers.
  • Use doubao-seedance-1-5-pro-251215 as the current production example for this route.
  • The same current model can handle both text-only and image-guided requests through the unified content array.

​
Task flow

1

Build the content array

Put the main text instruction first, then add image items only when you need first-frame or last-frame guidance.
2

Submit the task

Save the returned task id. If you want push delivery, set callback_url when you create the task.
3

Poll until output arrives

Call ByteDance Video Query until the task finishes and a content.video_url is present.

​
Known wrapper behavior

Volcengine and Seedance document this family as an asynchronous create-then-query workflow. CometAPI keeps that same task pattern but gives you one base URL and one API key.
During live validation on 2026-03-12, this route was reachable but still returned a server-side JSON parse error for a standard JSON sample. Treat the OpenAPI examples here as the practical wrapper reference, and validate your payload early in development.

Authorizations

​
Authorization
string
header
required

Bearer token authentication. Use your CometAPI key.

Body

application/json
​
model
string
required

ByteDance Seedance video model id. Choose a current model from the Models page. A current production example is doubao-seedance-1-5-pro-251215, which supports both text-only and image-guided requests through the same content array.

Example:

"doubao-seedance-1-5-pro-251215"

​
content
object[]
required

Ordered input items for the task. Put the main text instruction first. Add image items when you need first-frame or last-frame guidance.

​
callback_url
string

Optional webhook URL for task status notifications.

Response

200 - application/json

Task accepted.

​
id
string
required

Task id for later polling.