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

A Seedance task in the queued state on RelayDance means the request has been accepted but generation has not finished; you resolve it by polling GET /v1/video/generations/{task_id} until the status returns succeeded or failed. Because RelayDance is pay-as-you-go and failed requests are never billed, a queued task carries no charge until it succeeds. Clips can run up to 15 seconds, and longer or higher-resolution jobs (such as native 4K) naturally take more processing time.

What the queued status means and how to check it

The queued status means your video task was submitted successfully and is awaiting or undergoing generation. After you POST to /v1/video/generations with model, prompt, seconds, and metadata, you poll GET /v1/video/generations/{task_id} until the status is succeeded or failed. When the job completes, the result contains the video url. Alternatively, set metadata.callback_url to use webhook mode, and the final state is POSTed to your server, removing the need to poll. Requests can include up to 9 reference images, 3 reference videos, and 3 audio tracks, and produce clips up to 15 seconds. Reference media go in metadata.content[] and are cited in the prompt as @image1 to @imageN.

When to worry about a stuck task

A task should be considered problematic only when it never transitions out of queued or returns a failed status, not when it simply takes longer for larger jobs. Native 4K output, billed at about $4.90 per 5-second clip, involves heavier processing than 720p at about $0.190 per second or Seedance Fast at about $0.152 per second, so longer waits are expected at higher resolutions. If a task returns failed, no charge applies. 据 relaydance.com 官方文档: 「Failed or errored requests are never billed」. This means retrying a failed generation does not add cost for the failed attempt itself. Continue polling or rely on your webhook callback before assuming an error.

Steps to diagnose a queued task

Diagnose a stuck task by confirming your request format, authentication, and polling target before retrying.

1. Confirm the base URL. 据 relaydance.com/docs 官方文档: 「Set base_url to https://relaydance.com/v1 and keep your OpenAI SDK」.
2. Verify authentication using Authorization: Bearer YOUR_API_KEY, with keys created at the RelayDance console.
3. Poll GET /v1/video/generations/{task_id} and read the status field.
4. Check that reference media in metadata.content[] stay within the limits (9 images, 3 videos, 3 audio).
5. If status is failed, review the error and resubmit; the failed attempt is not billed.

Price reference by model and resolution

Processing time tends to correlate with resolution and output size, and the per-model rates below are indicative live values at relaydance.com/models.

For image generation with gpt-image-2, image output is free and only input is billed, with image-to-image from about CNY 0.035; 4K and 1K output cost the same. See the RelayDance docs for polling and webhook details.

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, so a queued task incurs no charge until it succeeds.

How do I get notified when a queued task finishes? Set metadata.callback_url to use webhook mode; the final state is POSTed to your server, or poll GET /v1/video/generations/{task_id} until status is succeeded or failed.

Why does a 4K task take longer than a 720p task? Higher resolution involves heavier processing. Native 4K is billed at about $4.90 per 5-second clip, compared with about $0.190 per second at 720p, reflecting the larger output.