ByteDanceSeed Speech
Seed ASR 1.0
ASR 1.0 remains a named resource in the audio-file API. The legacy bidirectional streaming mode supports this generation.
Technical guide
The streaming guide distinguishes the legacy endpoint from the optimized endpoint. Its language and recognition options are mode-specific.Seed ASR streaming API
Practical use
Practical guidance. When maintaining an older integration, verify its endpoint and resource ID together. Changing only one can alter which features the service accepts.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Access | API |
| Model identifiers | volc.bigasr.auc |
| File resource | volc.bigasr.auc.Seed ASR audio-file API |
Related models
- Seed ASR 2.0
ByteDance ยท Seed Speech
Speech recognition model with batch transcription for recordings shorter than five hours.
Launch date not confirmedAPI