Chatterbox TTS
Open text-to-speech models with zero-shot voice cloning in 23+ languages
What it is
Chatterbox is a family of open text-to-speech models from Resemble AI, ranging from a 110M CPU-friendly Nano to a 500M multilingual model, installable as a Python package. It clones a voice from a short reference clip, supports paralinguistic tags like [laugh] and [cough], and offers controls for expressiveness. Developers building voice agents, narration tools, or localized apps can run it locally on GPU or CPU.