AI · GitHub repo
qwen3_tts_rs
Rust-based inference CLI and API server for Qwen3 TTS with voice cloning
qwen3_tts_rs is a native Rust implementation for Qwen3 text-to-speech model inference. It packages TTS, voice cloning, and an OpenAI-compatible audio API server into standalone CLI binaries without requiring a Python runtime. It supports hardware acceleration using Apple Silicon MLX via Metal GPU as well as libtorch with NVIDIA CUDA.
What it's for
- Synthesize speech locally from text using CLI commands or agent workflows
- Clone target voices from short reference audio clips using the Qwen3 base model
- Host a local OpenAI-compatible /v1/audio/speech server for AI agents
- Integrate voice generation skills into autonomous agent frameworks like OpenClaw
Features
- Self-contained binary builds without Python runtime dependencies
- Apple Silicon acceleration via native MLX and Metal GPU
- Linux and NVIDIA CUDA acceleration via libtorch
- Voice cloning mode using reference audio and transcript prompts
- Predefined named speaker voices including Vivian, Serena, and Ryan
- Voice style and emotion control via natural language instructions on supported models
- OpenAI-compatible HTTP API server (/v1/audio/speech)
- Audio encoding support across multiple formats including WAV, MP3, FLAC, and OGG Opus
Pricing
Open source
Details
- Type
- GitHub repo
- Category
- AI
- Topic
- Text to Speech
- Open source
- Yes
- Platforms
- Command line, Linux, macOS, API, Self-hosted
- For
- Developers and AI engineers running local text-to-speech and voice cloning pipelines for bots and agents.
- Alternative to
- ElevenLabs, ChatTTS, Bark, Piper
- Tags
- tts, rust, voice cloning, speech synthesis, mlx, cuda, ai agents
Why it was saved
- @juntao: No external dependencies, cross-platform GPU acceleration (libtorch + NVIDIA on Linux, MLX + Apple GPU/NPU on macOS), voice cloning from 3-second clips, and voice instructions for emotion/style. Designed as a CLI for AI agents and humans.
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