Skip to content

elevenlabs.md

Latest commit

 

History

History
50 lines (36 loc) · 1.35 KB

elevenlabs.md

File metadata and controls

50 lines (36 loc) · 1.35 KB
Raw

ElevenLabs Engine

ElevenlabsEngine uses the ElevenLabs streaming text-to-speech API. It is a cloud engine and requires an ElevenLabs API key.

Install

pip install "realtimetts[elevenlabs]"

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

$env:ELEVENLABS_API_KEY = "..."

Install mpv for local playback of the MPEG stream.

Minimal Use

from RealtimeTTS import TextToAudioStream, ElevenlabsEngine


if __name__ == "__main__":
    engine = ElevenlabsEngine(voice="Nicole")
    stream = TextToAudioStream(engine)
    stream.feed("Hello from ElevenLabs.")
    stream.play()

Source Notes

  • Constructor defaults include voice="Nicole", id="piTKgcLEGmPE4e6mEKli", category="premade", and model="eleven_multilingual_v2".
  • get_voices() calls the ElevenLabs API and returns ElevenlabsVoice objects.
  • The current source comments note that newer ElevenLabs stream endpoints do not accept clarity, stability, or style values per request. The constructor keeps those fields for compatibility.
  • Output reports pyaudio.paCustomFormat, -1, -1, so compressed playback needs the external player path.

Troubleshooting

  • Missing key errors can be fixed by passing api_key or setting ELEVENLABS_API_KEY.
  • If a voice name is not found, pass the explicit voice ID in the constructor.