chatterbox: SoTA open-source TTS

Chatterbox, an open-source text-to-speech model from Resemble AI, offers production-grade quality under an MIT license. It outperforms closed-source systems and features emotion exaggeration control. With a 0.5B Llama backbone and 0.5M hours of training data, Chatterbox delivers stable, high-quality speech synthesis. The model includes built-in watermarking and can be easily installed for various applications. Resemble AI also offers a competitively priced TTS service for higher accuracy and scalability.