After you create a Seedance 2.0 video generation task, use this endpoint to check its status and retrieve the download URL once generation is complete. Pass theDocumentation Index
Fetch the complete documentation index at: https://docs.xingchaoyiqing.com/llms.txt
Use this file to discover all available pages before exploring further.
id returned by the Create Video Task endpoint as the path parameter, and keep polling until status is "completed" or "failed".
Video generation is asynchronous. Fast model variants (
seedance_2_0_fast, seedance_2_0_fast_pro) typically complete sooner and can be polled at shorter intervals.Endpoint
Path Parameters
The unique task identifier returned by the Create Video Task endpoint.Example:
video_fd35ee52-2a98-44a6-b930-29a88ce9b8fdResponse Fields
Unique identifier of the video generation task.
Object type. Always
"video".Unix timestamp (seconds) of when the task was created.
The Seedance 2.0 model used for generation. One of:
seedance_2_0, seedance_2_0_fast, seedance_2_0_pro, seedance_2_0_fast_pro.Current task status:
queued— task is in the queue, waiting to be processedprocessing— the model is actively generating the videocompleted— generation succeeded;video_urlis populatedfailed— generation failed; seeerrorfor details
A direct download URL for the generated MP4 video. Only present when
status is "completed". Valid for 1 day — download and store the video promptly.The actual output duration of the video in seconds. Only present when
status is "completed".The total number of tokens consumed by this generation task. Only present when
status is "completed". Use this value for cost tracking and quota management.Error message describing why the task failed. Only present when
status is "failed".Status Flow
| Step | Status | Meaning |
|---|---|---|
| 1 | queued | Task created and added to the processing queue |
| 2 | processing | Seedance 2.0 model is generating the video |
| 3 | completed | Video is ready; retrieve it via video_url |
| — | failed | An error occurred; inspect the error field |
Polling Recommendations
| Model | Recommended Poll Interval | Estimated Generation Time |
|---|---|---|
seedance_2_0 | every 20–40 s | 2–5 min |
seedance_2_0_fast | every 10–20 s | 1–3 min |
seedance_2_0_pro | every 20–40 s | 3–6 min |
seedance_2_0_fast_pro | every 10–20 s | 2–4 min |
- Stop polling immediately when
statusis"completed"or"failed". - Avoid polling more frequently than every 10 seconds to prevent rate limiting.
- Video duration: longer durations take more time
- Model type: fast variants generate more quickly
- Reference media: adding multi-modal references (images, videos, audio) may increase processing time
- System load: peak usage periods may result in longer queue times
Video URL Notes
- Format: MP4
- Audio: present if
generateAudiowastrueduring task creation; silent iffalse - Retention: 1 day