Skip to main content

Snorbyte TTS — CURL Recipe

POST https://api.snorbyte.com/tts

All parameters are query params (on the URL), not JSON.
Key params:

  • api_key (required,string) - Your API Key
  • utterance (required,string) - Text to speak (URL-encode it)
  • speaker_id (string, default: "49") - Use either speaker_name or speaker_id
  • speaker_name (string, case-insensitive) - Use either speaker_name or speaker_id
  • tone (required, string) - Only available for few speakers, for other set it to ""
  • temperature (optional, float, default: "0.0", can affect latency) - Model Temperature Tuning
  • top_p (optional, float, default: "1.0", can affect latency) - Model Top_p Tuning
  • repetition_penalty (optional, default: "1.05", can affect latency) - Model repetition_penalty Tuning
  • speed (optional, float, default: 1.00, can affect latency) - Speed up/down parameter
  • denoise (optional, bool, default: true) - Denoising for removing background noise if any
  • stream (optional, bool, default: true) - Streaming Audio for Low latency First Audio (Sub 250ms)
  • stream_format *(string, default: "mp3") — one of mp3|wav|pcm - ata_format of output bytes

Tip: Add --data '' to your curl call so it’s a true POST with empty body while parameters stay in the URL.

Quick MP3 Stream​

curl.exe -sS -L -X POST "http://api.snorbyte.com/tts?api_key=<YOUR_API_KEY>&utterance=Hello%20from%20Snorbyte%21&speaker_id=49&tone=&stream=true&stream_format=mp3" --data "" | ffplay -nodisp -autoexit -