> ## Documentation Index
> Fetch the complete documentation index at: https://docs.anyfast.ai/llms.txt
> Use this file to discover all available pages before exploring further.
# Seedance 2.0 Task Query
> Query the status and result of ByteDance Seedance video generation tasks.
After creating a Seedance 2.0 video generation task, use this endpoint to poll its status and retrieve the result when completed.
## Endpoint
```
GET /v1/video/generations/{task_id}
```
## Quick example
```bash cURL theme={null}
curl https://www.anyfast.ai/v1/video/generations/asyntask_h81efT0qwtLnzyt6VENmCZ6VXSfNxE2E \
-H "Authorization: Bearer YOUR_API_KEY"
```
```python Python theme={null}
import requests, time
task_id = "asyntask_h81efT0qwtLnzyt6VENmCZ6VXSfNxE2E"
while True:
r = requests.get(
f"https://www.anyfast.ai/v1/video/generations/{task_id}",
headers={"Authorization": "Bearer YOUR_API_KEY"}
).json()
status = r["data"]["status"]
if status == "SUCCESS":
print("Video URL:", r["data"]["result_url"])
break
if status == "FAILURE":
print("Failed:", r["data"]["fail_reason"])
break
time.sleep(15)
```
## Status values
| `data.status` | Meaning | Action |
| ------------- | ----------------------------------------- | -------------------------------------------- |
| `NOT_START` | Task accepted, not yet queued | Keep polling |
| `QUEUED` | Queued | Keep polling |
| `IN_PROGRESS` | Processing (`progress` advances from 50%) | Keep polling, \~15s interval |
| `SUCCESS` | Done | Read `data.result_url` |
| `FAILURE` | Failed | Read `data.fail_reason`; quota auto-refunded |
Total generation time is typically 1–5 minutes. The `result_url` is valid for **24 hours** — download or transfer it immediately.
## Top-level response fields
| Field | Type | Description |
| -------------------------- | ------ | ---------------------------------------------------------------------------------------------------- |
| `data.task_id` | string | Platform task ID (format `asyntask_xxx`) |
| `data.action` | string | Sub-type: `generate`, `referenceGenerate`, `firstTailGenerate`, `omniGenerate` |
| `data.status` | string | Task status (see table above) |
| `data.result_url` | string | Pre-signed download URL, **valid 24 hours**. Omitted on failure |
| `data.original_result_url` | string | **Ultra tasks only.** Upstream original (pre-enhancement) URL |
| `data.fail_reason` | string | Error message on failure. *Backward compat:* on success, copied from `result_url` for legacy clients |
| `data.submit_time` | int64 | Submit timestamp (Unix seconds) |
| `data.start_time` | int64 | Upstream start timestamp; `0` if not started |
| `data.finish_time` | int64 | Finish timestamp |
| `data.progress` | string | Progress text, e.g. `"50%"`, `"100%"` |
| `data.request_id` | string | Request trace ID |
| `data.data` | object | Upstream raw output (see below) |
## `data.data` — upstream raw output
| Field | Description |
| ------------------------------- | --------------------------------------------------------------------------------------------------------------------------------------------- |
| `content.video_url` | Generated video URL (same as outer `result_url` for non-ultra) |
| `id` | Upstream task ID |
| `model` | Upstream model ID |
| `duration` | Video duration (seconds) |
| `ratio` | Aspect ratio |
| `resolution` | Upstream raw resolution. **For ultra tasks** this is the lower upstream resolution; the final user-facing resolution is in `super_resolution` |
| `super_resolution` | **Ultra only.** Final resolution after super-resolution (e.g. `720p`, `2k`) |
| `framespersecond` | FPS |
| `generate_audio` | Whether audio was generated |
| `seed` | Random seed used |
| `status` | Upstream status (`succeeded`, `failed`, `running`, ...) |
| `usage.completion_tokens` | Upstream raw completion tokens |
| `usage.total_tokens` | Upstream raw total tokens |
| `usage.super_resolution_tokens` | **Ultra only.** Tokens after super-resolution coefficient — actual billed tokens |
## Success response (standard)
```json theme={null}
{
"code": "success",
"message": "",
"data": {
"task_id": "asyntask_h81efT0qwtLnzyt6VENmCZ6VXSfNxE2E",
"action": "generate",
"status": "SUCCESS",
"result_url": "https://ark-acg-cn-beijing.tos-cn-beijing.volces.com/...mp4?...",
"fail_reason": "https://ark-acg-cn-beijing.tos-cn-beijing.volces.com/...mp4?...",
"submit_time": 1777288506,
"start_time": 1777288508,
"finish_time": 1777288800,
"progress": "100%",
"request_id": "20260427111505260254000NBdvZFoP",
"data": {
"content": {"video_url": "https://ark-acg-cn-beijing.tos-cn-beijing.volces.com/...mp4?..."},
"id": "cgt-20260427191505-9j2q7",
"model": "doubao-seedance-2-0-260128",
"duration": 5,
"ratio": "16:9",
"resolution": "720p",
"framespersecond": 24,
"generate_audio": true,
"seed": 4304,
"status": "succeeded",
"usage": {"completion_tokens": 108900, "total_tokens": 108900}
}
}
}
```
## Success response (ultra / super-resolution)
Ultra models (e.g. `seedance-turbo`) use a two-stage pipeline (upstream low-res → MediaKit super-resolution). Extra fields:
```json theme={null}
{
"code": "success",
"message": "",
"data": {
"task_id": "asyntask_2cEAvaCpERzvSo7wfCCPZmNlGx8cjnu4",
"action": "generate",
"status": "SUCCESS",
"result_url": "https://...vod.cn-north-1.volcvideo.com/...?preview=1&auth_key=...",
"original_result_url": "https://ark-acg-cn-beijing.tos-cn-beijing.volces.com/...mp4?...",
"fail_reason": "https://...vod.cn-north-1.volcvideo.com/...?preview=1&auth_key=...",
"submit_time": 1777288507,
"start_time": 1777288509,
"finish_time": 1777288834,
"progress": "100%",
"request_id": "20260427111507187625000wErxKeB6",
"data": {
"content": {"video_url": "https://ark-acg-cn-beijing.tos-cn-beijing.volces.com/...mp4?..."},
"id": "cgt-20260427191507-f5n97",
"duration": 5,
"ratio": "16:9",
"resolution": "480p",
"super_resolution": "720p",
"usage": {
"completion_tokens": 50638,
"total_tokens": 50638,
"super_resolution_tokens": 114441
}
}
}
}
```
> Note: `data.resolution = "480p"` is the upstream raw value (the system downsizes the user's 720p request to 480p upstream to save cost). `data.super_resolution = "720p"` is the user-facing final resolution.
## Failure response
```json theme={null}
{
"code": "success",
"message": "",
"data": {
"task_id": "asyntask_xxx",
"action": "generate",
"status": "FAILURE",
"fail_reason": "task failed, code: InvalidParameter, message: img_url must be set for image to video method",
"submit_time": 1775802896,
"start_time": 0,
"finish_time": 1775802902,
"progress": "100%",
"request_id": "..."
}
}
```
The platform **automatically refunds** any pre-deducted quota for failed tasks — no caller action needed.
View the interactive API playground for Seedance 2.0 Task Query.