-
Notifications
You must be signed in to change notification settings - Fork 23
Home
Welcome to the official documentation for the Chatterbox TTS API!
Chatterbox TTS API is a local, OpenAI-compatible text-to-speech service built with FastAPI. It enables you to generate high-quality, voice-cloned speech from text, with real-time streaming, voice library management, and rich configuration options. Use it anywhere OpenAI's TTS API is supported—locally, privately, and with your own voices.
- Getting Started
- API Reference
- Deployment & Docker
- Voice Upload & Library
- Streaming & Advanced Usage
- Migration & Upgrades
- Troubleshooting & FAQ
- Contributing
- OpenAI-Compatible: Drop-in for TTS endpoints, works with Open WebUI, AnythingLLM, and more.
- Voice Cloning: Use your own samples for custom voices.
- Real-Time Streaming: Get speech audio as it’s generated for fast feedback.
- Persistent Voice Library: Upload, manage, and select voices by name.
- FastAPI & Async: Modern Python stack, automatic documentation, and high concurrency.
- Docker & Native: Run anywhere, GPU/CPU auto-detection, optimized builds.
- Open Source, Local, and Private: No cloud required, your voices stay yours.
This wiki is organized so you can find the right info fast—whether you’re a user, developer, or contributor.
Install, configure, and launch the API in minutes. Tips for both pip and Docker.
All endpoints, parameters, request/response formats, and example calls.
Compose files, GPU/CPU variants, and advanced deployment strategies.
How to upload, manage, persist, and use custom voices.
Real-time streaming, chunking strategies, endpoint aliasing, and performance tuning.
Guides for moving from Flask to FastAPI, pip to uv, and keeping dependencies up-to-date.
Solutions to common problems, error explanations, and where to get help.
8. Contributing
How to get involved, code style, and submitting PRs or issues.
For anything missing or unclear, open an issue or suggest a wiki edit.