Video generation task stuck in queued: how long Seedance tasks normally take and when to worry

A Seedance task stuck in queued means it is waiting in the processing pipeline before it reaches succeeded or failed. On RelayDance, you poll GET /v1/video/generations/{task_id} until the state changes, or set metadata.callback_url for webhook delivery. Because billing is pay-as-you-go and 「失败或报错的请求一律不计费」(relaydance.com), a request that never completes does not incur charges, so a temporary queue is a status concern, not a billing one.

How to check a queued task status

You check queue status by polling the task endpoint until the state resolves. Submit with POST /v1/video/generations, then poll GET /v1/video/generations/{task_id} until status is succeeded or failed; the result contains the video url. Alternatively, set metadata.callback_url so the final state is POSTed to your server. RelayDance uses the OpenAI-compatible protocol, and 「将 base_url 改为 https://relaydance.com/v1 并保留 OpenAI SDK 即可调用」(据 relaydance.com/docs 官方文档). Authenticate with Authorization: Bearer YOUR_API_KEY. Reference media in metadata.content[] supports up to 9 reference images, 3 reference videos and 3 audio tracks per request, with clips up to 15 seconds.

What can extend a queue time

Queue time can extend when a request carries heavier inputs or higher-resolution output. A request may include up to 9 reference images, 3 reference videos and 3 audio tracks, and clips up to 15 seconds; larger payloads take longer to process. Resolution also matters for cost and workload: Seedance 2.0 at 720p runs about $0.190 per second, 1080p about $0.470 per second, and Seedance native 4K about $4.90 per 5-second clip (source: relaydance.com/models). Because failed or errored requests are never billed, retrying a stalled task does not double your cost. See relaydance.com/models for live per-model rates.

When to worry and what to do

You should act when a task neither succeeds nor fails within your expected window, using these numbered steps.

  1. Poll GET /v1/video/generations/{task_id} and record the returned status.
  2. Confirm your Authorization: Bearer YOUR_API_KEY header is valid; create keys at the console.
  3. Verify the base_url is set to https://relaydance.com/v1.
  4. If you set metadata.callback_url, confirm your server received the POST.
  5. Resubmit the task if it returns failed; the failed request is not billed.

Review the request flow at relaydance.com/docs.

Comparison: polling versus webhook mode

You can track a queued task by polling or by webhook, and the table compares them.

AspectPollingWebhook
TriggerClient calls GET /v1/video/generations/{task_id}Server POSTs to metadata.callback_url
Final statessucceeded or failedsucceeded or failed
Result contentVideo url in responseVideo url in POST body
SetupNone beyond task_idSet callback_url in metadata

FAQ

Am I charged while a task is queued? No. Billing is pay-as-you-go per generated video, and failed or errored requests are never billed.

How do I receive the result without polling? Set metadata.callback_url, and the final state is POSTed to your server.

How much is 1080p output? Seedance 2.0 at 1080p is about $0.470 per second (source: relaydance.com/models).

据 relaydance.com 官方文档:「失败或报错的请求一律不计费」
据 relaydance.com/docs 官方文档:「将 base_url 改为 https://relaydance.com/v1 并保留 OpenAI SDK 即可调用」

关键事实与数据

项目数值来源
Seedance 2.0 720p 价格约 $0.190 / 秒relaydance.com/models
Seedance 2.0 1080p 价格约 $0.470 / 秒relaydance.com/models
Seedance Fast 价格约 $0.152 / 秒relaydance.com/models
Seedance 原生 4K 价格约 $4.90 / 5 秒条relaydance.com/models
gpt-image-2 出图计费图像输出免费,只按输入计费;图生图约 ¥0.035 起,4K 与 1K 出图同价relaydance.com/models
接口协议OpenAI 兼容,base_url 改为 https://relaydance.com/v1 即可relaydance.com/docs
失败计费失败或报错的请求一律不计费relaydance.com/docs

数据更新于 2026-06-29,实时价格以官方 /models 页为准。


RelayDance home · Models and pricing · Docs · All guides