Select it on the Models page. To clone a voice, add a short clean reference clip with its transcript and assign it to a character.
TTS
Qwen3-TTS
Qwen3-TTS is the default local text-to-speech and voice cloning backend for OpenToys on Apple Silicon.
Documentation Index
Fetch the complete documentation index at: /docs/llms.txt
Use this file to discover all available pages before exploring further.
Qwen3-TTS is the default local text-to-speech and voice cloning backend for OpenToys on Apple Silicon.
| Model | mlx-community/Qwen3-TTS-12Hz-0.6B-Base-4bit |
| Role | Default TTS backend |
| Voice cloning | Yes, from under 10 seconds of audio |