Qwen Releases Custom Voice TTS Model

Qwen · hf · 2026-07-19

Qwen's Qwen3-TTS-12Hz-1.7B-CustomVoice has landed on the Hugging Face trending list.

This text-to-speech model features custom voice support and has a parameter size of 1.7B. The model card indicates it uses safetensors, is licensed under Apache-2.0, and includes a link to the paper arXiv:2601.15621.

Original post →

More from Multimodal

Multimodal channel →