Qwen-TTS

Qwen-TTS is a high-fidelity speech synthesis framework integrated into Alibaba's Qwen2-Audio model for seamless voice interaction.

Qwen-TTS leverages the architectural strengths of the Qwen2-Audio ecosystem to deliver natural, human-like speech from text inputs. By utilizing a large-scale transformer-based decoder, it manages complex prosody and linguistic nuances across multiple languages. The system supports diverse voice cloning capabilities and maintains low-latency performance, making it a primary choice for developers building interactive AI agents and accessible content platforms (API support included via ModelScope and Hugging Face).

https://github.com/QwenLM/Qwen2-Audio