Chatterbox TTS Server
Self-hosted text-to-speech server runs Chatterbox models on CPU or GPU, with voice cloning and an OpenAI-compatible API.
Alternative to:
Chatterbox TTS Server turns text into speech on your own computer or server. It's for people making narration and audiobooks, or adding speech to apps through an API.
- Choice of speech models: Original Chatterbox produces English speech with emotion control, Multilingual supports speech in multiple languages, and Turbo generates speech faster and accepts tags for laughter, coughing and chuckling.
- Voices and cloning: Choose a built-in voice or provide a WAV or MP3 recording to generate speech in a similar voice. A fixed generation seed can help keep output consistent across requests.
- Long-form narration: The server splits long text at sentence boundaries and joins the audio into one file, including for book-length input.
- Browser and API access: Generate speech through the web interface or an OpenAI-compatible API, and switch between Chatterbox models without restarting the server.
The server can use NVIDIA, AMD or Apple Silicon GPUs, with CPU support as well. Docker deployment is available, and the project has an MIT license. An optional Google Colab demo runs in Google's cloud rather than on your hardware.
Logotyper
