API Reference

Authentication

All Shunya TTS endpoints require a valid API key. Authentication differs slightly between HTTP and WebSocket transports.

HTTP (Bearer Token)

For batch synthesis, pass your API key in the Authorization header:

terminal
curl -X POST https://tts.shunyalabs.ai/v1/audio/speech \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"zero-indic","voice":"Nisha","input":"Hello world"}'

WebSocket (Query Param)

For streaming synthesis, pass the API key as a query parameter when opening the WebSocket connection:

python
import websockets

ws_url = "wss://tts.shunyalabs.ai/ws?api_key=YOUR_API_KEY"

async with websockets.connect(ws_url) as ws:
    # Send config and text frames...
    pass

Authentication Errors

  • 401 Unauthorized -- Missing or invalid API key.
  • WebSocket connections with an invalid key are closed immediately with code 4001.
PreviousOverview & Base URLsNext Batch Synthesis