Test the Seedance API with curl: minimal working requests for every endpoint

To test the Seedance API with curl, point your requests at the OpenAI-compatible base URL https://relaydance.com/v1, add an Authorization: Bearer YOUR_API_KEY header, then POST to /v1/video/generations to submit a task and GET /v1/video/generations/{task_id} to poll until the status is succeeded or failed. Images use /v1/images/generations. Seedance 2.0 720p runs at about $0.190 per second, and failed requests cost nothing.

Submit a video task with curl

Send a POST request to /v1/video/generations with the model, prompt, and seconds fields to start a Seedance job.

1. Set your endpoint: POST https://relaydance.com/v1/video/generations
2. Add the header Authorization: Bearer YOUR_API_KEY (create keys at the console).
3. Send a JSON body with model, prompt, and seconds.
4. Optionally add metadata fields: ratio, resolution, generate_audio, and callback_url.

Reference media go in metadata.content[] and are cited in the prompt as @image1 to @imageN. A single request accepts up to 9 reference images, 3 reference videos, and 3 audio tracks, with clips up to 15 seconds. As the official relaydance.com/docs docs state, "Set base_url to https://relaydance.com/v1 and keep your OpenAI SDK", so the same curl pattern works across models.

Poll for the result or use a webhook

After submitting, poll GET /v1/video/generations/{task_id} until the status returns succeeded or failed.

When the status is succeeded, the response contains the video URL. Repeat the GET call at your own interval until you reach a terminal state. If you prefer not to poll, use webhook mode: set metadata.callback_url in the original submission, and the final state is POSTed to your server automatically. Both approaches use the same task lifecycle. Because billing is pay-as-you-go per generated video, only completed jobs are charged. According to the official relaydance.com docs, "Failed or errored requests are never billed", so retries during testing carry no cost for errored tasks.

Generate images with curl

Image generation uses the same auth pattern against the /v1/images/generations endpoint.

Send a POST request to https://relaydance.com/v1/images/generations with your Bearer token and a JSON body. For gpt-image-2, image output is free and only input is billed, so image-to-image starts at about CNY 0.035. Note that 4K and 1K output cost the same under this model. This billing structure differs from per-second video pricing, where resolution changes the rate. See https://relaydance.com/models for live per-model rates and confirm current numbers before running large test batches.

Compare Seedance pricing before testing

Seedance rates vary by model and resolution, so pick the tier that matches your test.

For low-cost validation, Seedance Fast at about $0.152 per second is the lightest option. Payments accept USDT and Stripe card, and access does not require a BytePlus enterprise account or KYC. Full details are documented at https://relaydance.com/docs.

FAQ

Do I need an OpenAI SDK to call Seedance? No. The API is OpenAI-compatible, so you can use plain curl or an existing OpenAI SDK by setting the base_url to https://relaydance.com/v1.

Am I charged if a request fails? No. Per relaydance.com, failed or errored requests are never billed.

How do I get the finished video URL? Poll GET /v1/video/generations/{task_id} until the status is succeeded, then read the video URL from the response, or set metadata.callback_url to receive the final state via webhook.