ChatTTS

ChatTTS

ChatTTS is a powerful generative speech model designed specifically for dialogue applications. It enables natural and expressive speech synthesis across multiple speakers, making it ideal for interactive conversations.

Key Features

• Conversational TTS: Optimized for dialogue-based tasks, providing natural-sounding speech.
• Fine-grained Control: Predicts and controls prosodic features like laughter and pauses for more expressive speech.
• Multi-lingual Support: Trained with over 100,000 hours of audio data in Chinese and English, supporting mixed-language input.

Benefits

• Enhanced Prosody: Surpasses most open-source TTS models in terms of prosody.
• Open-source: The model is available for academic and research purposes, built to support further developments in speech synthesis.

Highlights

• Streaming audio generation capability.
• Active community contributions and ongoing development under the AGPLv3+ license.