Qwen3-TTS
Voice cloning requires the Base model. Set TTS_MODEL_NAME=Qwen/Qwen3-TTS-12Hz-1.7B-Base and restart.
Upload a reference audio clip (5-30 seconds of clear speech) and its transcript to create a reusable voice profile. Once saved, the voice appears in the TTS voice dropdown.
Click or drag & drop an audio file
WAV, MP3, FLAC, etc.
Your Voice Profiles
No voice profiles yet. Go to Voice Cloning to create one.