Global

VoxCPM is a tokenizer-free Text-to-Speech system that directly generates continuous speech representations via an end-to-end diffusion autoregressive architecture, bypassing discrete tokenization to achieve highly natural and expressive synthesis.

Chatterbox is a family of state-of-the-art, open-source text-to-speech models by Resemble AI.

Open-source State-of-the-art TTS model which runs on a CPU 😻

5B CosyVoice-300M CosyVoice-300M-SFT CosyVoice-300M-Instruct model and CosyVoice-ttsfrd resource.

🐸TTS is a library for advanced Text-to-Speech generation.