chatterbox
by resemble-ai
Chatterbox is a family of open-source text-to-speech models from Resemble AI that turn written text into spoken audio and can copy a voice from a short reference clip. It is for adding voiceovers, voice agents or multilingual speech to apps and content.
- Stars
- 27.3k
- Licence
- MIT
- Runs on
- A gpu machine, with a GPU
- Setup
- Moderate
- Language
- Python
- Age
- 552 days
What you can do with it
- Generate spoken audio from text in a cloned voice using a roughly 10-second reference clip
- Produce speech in 23+ languages with the multilingual model
- Add expressions such as [laugh] or [cough] to English narration and voice agent replies
Make money with it
Offer voiceover, audiobook or localised video narration services to clients, using the models to produce the audio quickly.
An idea, not financial advice. Check the licence before charging for anything built on it.
Ways a business could use it
- Create voiceovers for videos, adverts and training material without paying per-minute fees to a voice service
- Build a phone or website voice assistant that answers customers in a consistent brand voice, including in other languages
Can you use it commercially?
Free to use, change and sell, including in paid products. Keep the copyright notice.
Not legal advice. Read the project's licence file before building a product on it.
Getting it running
Install with "pip install chatterbox-tts" (Python 3.11 recommended) and run the models from a short Python script, ideally on a CUDA GPU.
Follow the official instructions in the project README.
Things to check
- Voice cloning can be misused for impersonation, so get consent from anyone whose voice you copy
- Best performance needs a GPU, though the small Nano model can run on CPU
- Tested mainly on Python 3.11 and Debian, so Windows setup may need extra effort
Checked 27 Oct 2026. Summary drafted by AI from the project's README.
