This project is a text-to-speech model specifically designed for dialog scenarios, supporting both Chinese and English. It is capable of producing not only natural and smooth speech but also controlling and adding elements like laughter, pauses, and intonations. The ChatTTS model achieves excellent results in generating shorter pieces of speech, making the AI 'taste' almost imperceptible.