Breeze TTS 2, an open-weight text-to-speech model, has been released, achieving top rankings on the Artificial Analysis TTS leaderboard and surpassing proprietary systems. This model supports real-time interaction with ultra-low latency, achieving under 40ms time to first audio on an NVIDIA H100. It offers advanced features like voice cloning, voice design from natural language descriptions, and vocal event integration for expressive speech in both English and Chinese. AI
IMPACT Sets a new benchmark for open-weight TTS models, potentially driving innovation in real-time voice generation and voice cloning.
RANK_REASON Open-weight model release with benchmark performance claims. [lever_c_demoted from frontier_release: ic=1 ai=1.0]
Read on Hugging Face Trending Models →
- Artificial Analysis
- Breeze TTS 2
- CUDA
- Google Colab
- Kaggle
- Linux
- NVIDIA
- NVIDIA H100
- Python
- transformers
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →