Skip to content

cartesia.md

Latest commit

 

History

History
45 lines (31 loc) · 1.19 KB

cartesia.md

File metadata and controls

45 lines (31 loc) · 1.19 KB
Raw

Cartesia Engine

CartesiaEngine uses the Cartesia API over its WebSocket TTS connection.

Install

Install RealtimeTTS with the Cartesia extra:

pip install "realtimetts[cartesia]"

Set the API key in the environment or pass it to the constructor:

$env:CARTESIA_API_KEY = "..."

Minimal Use

from RealtimeTTS import TextToAudioStream, CartesiaEngine


if __name__ == "__main__":
    engine = CartesiaEngine(voice_id="your-cartesia-voice-id")
    stream = TextToAudioStream(engine)
    stream.feed("Hello from Cartesia.")
    stream.play()

Source Notes

  • Default model_id is sonic-3; valid model IDs in source include sonic-2, sonic-3, sonic-turbo, and sonic.
  • Default output is raw pcm_f32le at 44100 Hz. pcm_s16le is also supported.
  • language and add_timestamps are optional request fields.
  • CartesiaVoice(id, name=None, language=None) can be used for voice selection.
  • get_voices() fetches one page by default. Set fetch_all_voices=True for full pagination.

Troubleshooting

  • Missing Cartesia API key: pass api_key or set CARTESIA_API_KEY.
  • No voice set: pass voice_id, a CartesiaVoice, or a dict with id.