AI · Web app
Qwen3-TTS-12Hz-1.7B-CustomVoice
Open-weight text-to-speech model supporting instruction-driven style control across 10 languages
Qwen3-TTS-12Hz-1.7B-CustomVoice is an open-source text-to-speech model from Alibaba's Qwen team that allows fine-grained voice style and emotional direction via natural-language prompts. It supports 10 languages and multiple built-in speaker timbres with ultra-low latency streaming synthesis. It is useful for generating realistic dialogue and character voices without robotic cadence.
What it's for
- Synthesize expressive spoken audio directed by natural language emotion and tone prompts
- Generate multilingual character dialogue across English, Chinese, Japanese, and other languages
- Serve low-latency voice responses for real-time conversational agents and voice bots
Features
- Supports 10 major languages (English, Chinese, Japanese, Korean, German, French, Russian, Portuguese, Spanish, Italian)
- Instruction-driven acoustic control over pitch, speed, emotion, tone, and prosody
- 9 built-in speaker timbres covering varying genders, ages, and regional dialects
- Low-latency dual-track streaming generation capable of sub-100ms first-packet audio response
- Integration with Python via the qwen-tts package and vLLM-Omni offline inference
- Bundled local Gradio web UI demo executable from the command line
Pricing
Open source. Free and open source under Apache 2.0; commercial cloud API available via Alibaba DashScope
Details
- Type
- Web app
- Category
- AI
- Topic
- Text to Speech
- Open source
- Yes
- Platforms
- Linux, Windows, macOS, Code library, API, Command line
- For
- Developers, AI audio researchers, game sound designers, and creators needing open-weight expressive speech synthesis
- Alternative to
- ElevenLabs, F5-TTS, ChatTTS, OpenAI Audio API
- Tags
- tts, speech synthesis, voice cloning, audio generation, huggingface, qwen
Why it was saved
- @HuggingModels: Voice cloning that sounds actually real, usable in the browser and open source — plus the demo's voice-direction prompt template (gender/pitch/speed/volume/age/clarity/fluency/accent/texture/emotion/tone/personality) worth reusing.
Compare
Related
AI voice generator, voice cloning, and conversational agent platform
Ultra-low-latency streaming text-to-speech and instant voice cloning model.
Open-source 1.6B text-to-speech model generating multi-speaker dialogue in one pass.
Fast, low-cost text-to-speech API and studio for creators and developers
Converts spoken thoughts into structured, polished writing in customizable styles.
Official Python library for Cartesia's real-time text-to-speech API
Updated