Section 2
Pre-recorded Audio
Transcribe audio files, in-memory buffers, or remote URLs via HTTP POST. Receive a full structured transcript with optional NLP enrichment in a single response.
How it works
| Property | Value |
|---|---|
| Transport | HTTP POST — multipart/form-data |
| Endpoint | https://asr.shunyalabs.ai/v1/audio/transcriptions |
| Auth | Authorization: Bearer <API_KEY> |
| Required fields | model + one of: file or url |
| Model | zero-indic |
| NLP engine | Intent, summarization, sentiment, profanity, translation powered by Gemini |
When to use batch
- Call center recordings and voice notes
- Podcast, video, and media transcription
- Offline audio processing pipelines
- Any use case where real-time latency is not critical