Seedance Video Generation API

Video generation takes longer than a chat request, so the Tokuse endpoint returns a task ID instead of holding the connection. Submit the job, poll the task, then use the video once the status is succeeded.

Item	Value
Submit	POST /v1/video/generations
Poll	GET /v1/video/generations/{task_id}
Authentication	Authorization: Bearer $TOKUSE_API_KEY
Models	Seedance video models available to your account
Pricing	Tokuse pricing and the upstream provider console

Endpoint and authentication

Both the submit and poll calls use a Bearer token on https://tokuse.com. The submit call returns a task_id. The poll call takes that ID in the path and reports the task state.

Models and parameters

Use the Seedance model ID available to your account. Known working values include duration: 5, resolution: 480p or 720p, and ratio: "16:9". Treat those as a starting point, not a guarantee for every model.

Submit a task

curl https://tokuse.com/v1/video/generations \
  -H "Authorization: Bearer $TOKUSE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "your-seedance-model",
    "prompt": "A slow dolly shot through a data center aisle",
    "duration": 5,
    "resolution": "720p",
    "ratio": "16:9"
  }'

The submit response carries the task ID:

{
  "task_id": "task_7d91c2"
}

Poll the task

curl https://tokuse.com/v1/video/generations/task_7d91c2 \
  -H "Authorization: Bearer $TOKUSE_API_KEY"

A finished task looks like this:

{
  "task_id": "task_7d91c2",
  "status": "succeeded",
  "video_url": "https://example.com/generated/video.mp4"
}

Notes and limits

• Poll until status is succeeded. Add a delay between polls instead of a tight loop; the task runs asynchronously and repeated calls do not speed it up.
• The submit call returns a task ID, not a video URL. Store the ID if the client disconnects before the task finishes.
• Duration, resolution, and ratio affect generation time and cost. Test a short clip first when you switch models.

Billing

This page does not list unit prices. Video generation is billed per task, with the rate depending on duration and resolution. Check the Tokuse pricing page and the upstream provider console for current rates; gateway pricing is on /pricing.