Join the waitlist for NeuralVox - full-cast AI audiobooks!
An API for Chatterbox TTS.
This library consists of two components:
- A client library for interacting with the Chatterbox TTS API.
- The API server itself that wraps the Chatterbox TTS model.
pip install chatterbox-serverTo install with support for the API server, run:
pip install chatterbox-server[server]If you are getting CUDA issues, you can try the experimental fork of Chatterbox TTS:
pip install "chatterbox-tts @ git+https://github.com/fakerybakery/better-chatterbox@fix-cuda-issue"
Start the API server:
chatterbox-serverfrom chatterbox_api import ChatterboxAPI
api = ChatterboxAPI("http://localhost:5000")
# Make sure you have a server running!
# If you want to pass custom headers (e.g. for authentication if you are running behind a proxy/load balancer), you can do so like this:
# api = ChatterboxAPI("http://localhost:5000", headers={"Authorization": "Bearer <your-token>"})
response = api.synthesize(text="Hello, world!", audio_prompt="path/to/audio_prompt.wav")
# Save the response to a file
with open("output.wav", "wb") as f:
f.write(response.content)Copyright 2025 NeuralVox.
Licensed under the BSD-3-Clause license.