Skip to content

modelslab.md

Latest commit

 

History

History
52 lines (38 loc) · 1.49 KB

modelslab.md

File metadata and controls

52 lines (38 loc) · 1.49 KB
Raw

ModelsLab Engine

ModelsLabEngine uses the ModelsLab text-to-speech API. It exists under RealtimeTTS.engines, but the root RealtimeTTS package does not currently lazy-export it. Import it from RealtimeTTS.engines until that mismatch is resolved.

Install

Install RealtimeTTS with the ModelsLab extra. The source uses requests plus normal RealtimeTTS playback dependencies:

pip install "realtimetts[modelslab]"

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

$env:MODELSLAB_API_KEY = "..."

Minimal Use

from RealtimeTTS import TextToAudioStream
from RealtimeTTS.engines import ModelsLabEngine


if __name__ == "__main__":
    engine = ModelsLabEngine(voice="madison", language="american english")
    stream = TextToAudioStream(engine)
    stream.feed("Hello from ModelsLab.")
    stream.play()

Source Notes

  • Defaults are voice="madison", language="american english", speed=1.0, and emotion=False.
  • The source includes built-in voice IDs for American/British English, Spanish, French, German, Italian, Japanese, Hindi, Mandarin Chinese, and Brazilian Portuguese.
  • The API can return processing; the engine polls the returned fetch URL.
  • Output is MP3 and reports custom-format playback at 22050 Hz.

Troubleshooting

  • Missing key errors can be fixed by passing api_key or setting MODELSLAB_API_KEY.
  • Because the root package export is absent, prefer from RealtimeTTS.engines import ModelsLabEngine.