OpenAI SDK not reaching RelayDance: base_url and path configuration mistakes

If your OpenAI SDK is not reaching RelayDance, the cause is almost always a base_url or endpoint path mistake. Set base_url to https://relaydance.com/v1, keep your existing OpenAI SDK, and send requests to the correct paths (POST /v1/video/generations for video, POST /v1/images/generations for images). Authenticate with Authorization: Bearer YOUR_API_KEY. Because RelayDance is OpenAI-compatible, no SDK rewrite is required. See relaydance.com/docs for details.

Fix the base_url first

The most frequent connection failure is an incorrect base_url, so set it to https://relaydance.com/v1 before checking anything else. According to the official relaydance.com/docs, 「Set base_url to https://relaydance.com/v1 and keep your OpenAI SDK」. This means you do not swap SDKs or rewrite request logic: you point the same OpenAI client at RelayDance. Common errors include omitting the /v1 suffix, adding a trailing slash that duplicates path segments, or leaving a stale value that still targets another provider. Verify the value your client actually loads (environment variables can override code), then confirm authentication uses Authorization: Bearer YOUR_API_KEY, with keys created at the RelayDance console.

Use the correct endpoint paths

A 404 or unexpected response usually means the request path is wrong, not the host. Submit video tasks with POST /v1/video/generations, including model, prompt, and seconds, with optional metadata (ratio, resolution, generate_audio, callback_url, content[]). Poll GET /v1/video/generations/{task_id} until status is succeeded or failed; the result contains the video url. For images, use POST /v1/images/generations. Reference media go in metadata.content[] and are cited in the prompt as @image1 to @imageN. Per request you may include up to 9 reference images, 3 reference videos, and 3 audio tracks, with clips up to 15 seconds. Webhook mode: set metadata.callback_url so the final state is POSTed to your server.

Verify auth and billing before retrying

If requests are rejected, check the Authorization header before assuming a billing problem, because failures do not incur charges. According to the official relaydance.com docs, 「Failed or errored requests are never billed」. Billing is pay-as-you-go per generated video. Current rates include Seedance 2.0 at about $0.190 / second for 720p and about $0.470 / second for 1080p, Seedance Fast at about $0.152 / second, and Seedance native 4K at about $4.90 per 5-second clip (source: relaydance.com/models). Payments are accepted via USDT and Stripe card. Because failed or errored requests are never billed, you can safely retry a connection after fixing base_url or path issues.

Configuration checklist

Follow these numbered steps to confirm the SDK reaches RelayDance:

1. Set base_url to https://relaydance.com/v1 in your OpenAI client.
2. Set the header Authorization: Bearer YOUR_API_KEY using a key from the console.
3. Send video tasks to POST /v1/video/generations with model, prompt, and seconds.
4. Poll GET /v1/video/generations/{task_id} until status is succeeded or failed.
5. For images, send POST /v1/images/generations instead.

Endpoint reference

FAQ

Do I need to replace my OpenAI SDK? No. Keep your OpenAI SDK and set base_url to https://relaydance.com/v1.

Will failed connection attempts cost money? No. Failed or errored requests are never billed.

Which path do I use for video versus images? Use POST /v1/video/generations for video and POST /v1/images/generations for images.