Qwen/Qwen3-TTS-12Hz-1.7B-Base
🧠 AI ModelQwen
Open-source 1.7B parameter TTS model by Qwen with efficient 12Hz audio generation.
Qwen3-TTS-12Hz-1.7B-Base is a state-of-the-art text-to-speech model developed by Qwen, the AI research team behind the Qwen series. It uses a transformer architecture with 1.7 billion parameters to convert text into 12Hz low-frame-rate speech tokens. The innovative 12Hz tokenization reduces computational overhead while maintaining natural prosody and voice clarity, making it suitable for edge deployment and real-time systems. The model has been trained on diverse multilingual data and supports zero-shot voice cloning. With over 1.3 million downloads on Hugging Face, it has quickly become a popular choice for developers and researchers. The Apache 2.0 license allows free use, modification, and distribution.
💡Highlights
- ├─1.7B transformer parameters
- ├─12Hz low-frame-rate tokens
- └─Apache 2.0 license
🎯For
- ├─AI developers
- ├─voice interface engineers
- └─speech synthesis researchers