ChatTTS is a voice generation model designed for conversational scenarios, available on GitHub at 2noise/chattts. It caters to dialogue tasks for large language model assistants and conversational audio and video introductions, supporting both English and Chinese languages. With training on approximately 100,000 hours of data, ChatTTS delivers high-quality and natural speech synthesis. The project team plans to open-source a basic model trained with 40,000 hours of data to aid further research and development.

One of ChatTTS' key features is its multi-language support, including English and Chinese, overcoming language barriers. It has been extensively trained with about 10 million hours of data, resulting in top-notch voice synthesis. Ideal for dialog tasks, ChatTTS enhances interaction experiences in applications and services.