No sound in the generated video: checking audio options in Seedance 2.0 requests

If your Seedance 2.0 video has no sound, the most common cause is that metadata.generate_audio was not enabled when you submitted the task. RelayDance uses an OpenAI-compatible API, so audio is controlled per request through the metadata object on POST /v1/video/generations. Set generate_audio to true, resubmit, and poll the task until the status is succeeded. Because failed or errored requests are never billed, resubmitting carries no extra cost.

Why the audio is missing

The audio is missing because the generate_audio flag in metadata controls sound output and defaults to off unless you set it. When you call POST /v1/video/generations, you pass model, prompt, seconds, and a metadata object that holds ratio, resolution, generate_audio, callback_url, and content[]. If generate_audio is absent or false, the returned video has no audio track. Reference audio tracks (up to 3 per request) are separate from generated audio and also live in metadata.content[]. Verify both settings before you assume a defect in the model. Live per-model rates are listed at relaydance.com/models.

How to enable audio and resubmit

Enable audio by setting metadata.generate_audio to true, then submit and poll the task to completion.

1. Set base_url to https://relaydance.com/v1 and pass Authorization: Bearer YOUR_API_KEY.
2. Call POST /v1/video/generations with model, prompt, seconds, and metadata.generate_audio: true.
3. Poll GET /v1/video/generations/{task_id} until status is succeeded or failed.
4. Read the video url from the result and confirm the audio track plays.

According to the official relaydance.com/docs docs, 「Set base_url to https://relaydance.com/v1 and keep your OpenAI SDK」. According to the official relaydance.com docs, 「Failed or errored requests are never billed」, so retrying after a missing-audio task is free. Clips can run up to 15 seconds per request. See relaydance.com/docs.

Cost of resubmitting Seedance 2.0 tasks

Resubmitting a Seedance 2.0 task is billed pay-as-you-go per generated video, and only successful generations incur charges. Seedance 2.0 720p costs about $0.190 per second, and 1080p costs about $0.470 per second. Seedance Fast is about $0.152 per second, and Seedance native 4K is about $4.90 per 5-second clip. Because failed or errored requests are never billed, a retry with generate_audio enabled only costs you when it succeeds.

Using webhooks to confirm the final state

To confirm whether the finished video includes audio without polling, set metadata.callback_url so the final state is POSTed to your server. In webhook mode, RelayDance sends the succeeded or failed result, including the video url, to your endpoint. You can also cite reference media in the prompt as @image1 to @imageN, with limits of up to 9 reference images, 3 reference videos, and 3 audio tracks per request. Authentication uses Authorization: Bearer YOUR_API_KEY, and keys are created at the console. Payments are pay-as-you-go via USDT and Stripe card. Access is OpenAI-compatible with no BytePlus enterprise account or KYC required.

FAQ

Does Seedance 2.0 generate audio by default? No. You must set metadata.generate_audio to true on POST /v1/video/generations; otherwise the returned video has no audio track.

Will I be charged if my first task had no sound? You are charged only for successful generations. According to the official relaydance.com docs, 「Failed or errored requests are never billed」.

How do I know when the audio-enabled video is ready? Poll GET /v1/video/generations/{task_id} until status is succeeded, or set metadata.callback_url to receive the final state on your server.