Audio
Kokoro-FastAPI
Turn the beloved Kokoro voice model into an OpenAI-compatible speech API with one Docker command.
★ 5.5k starsApache-2.0
View guide →Alibaba's open voice models: expressive, streaming-capable speech with free-form voice design.
Qwen3-TTS is an open-source audio project by QwenLM (QwenLM/Qwen3-TTS on GitHub). It has 13,619 GitHub stars and 1,764 forks, is released under the Apache-2.0 license, and was last updated on GitHub on 17 March 2026.
Create an isolated Python 3.12 environment so the audio dependencies don't clash with your other projects.
One pip command pulls the runtime and lets you load any released Qwen3-TTS model.
Pick a model size for your hardware, feed it text, and get expressive speech — with streaming for realtime use.
Describe the voice you want in words, or provide a short sample to clone it, then reuse it across projects.
Natural intonation and emotion, well beyond robotic reader voices.
Audio starts flowing before the full text is processed — ideal for live agents and assistants.
Describe a voice in plain language (“warm, middle-aged narrator”) and get it.
Clone a voice from a short sample for consistent characters and branding.
Multiple released models so you can trade quality against speed and VRAM.
Permissive licensing for personal and commercial projects.
Make a fresh, isolated Python 3.12 environment.
conda create -n qwen3-tts python=3.12 -y
conda activate qwen3-ttsInstall the qwen-tts package from PyPI.
pip install -U qwen-ttsAdd FlashAttention 2 to cut GPU memory usage.
pip install -U flash-attn --no-build-isolationQwen3-TTS brings larger, more expressive models with streaming and voice design. Kokoro is lighter and faster to run. Choose expressive depth or lightweight speed.
A GPU is recommended for comfortable speeds; FlashAttention 2 substantially reduces the memory footprint.
Yes — streaming generation is a headline feature, built for realtime applications.
The Qwen team at Alibaba Cloud, the same group behind the popular Qwen language models.
Yes — it is released under the Apache-2.0 license.
Turn the beloved Kokoro voice model into an OpenAI-compatible speech API with one Docker command.
The simplest way to run open language models on your own machine.
Run and fine-tune LLMs and diffusion models locally — faster training, smaller memory footprint.
Chatterbox — Resemble AI's open-source TTS family: expressive text-to-speech and voice cloning, multilingual V3, low-latency Turbo and tiny Nano models, MIT
MOSS-TTS family: open text-to-speech models for voice cloning, dialogue, voice design, and real-time agents
VoxCPM2: tokenizer-free TTS for multilingual speech, creative voice design and true-to-life cloning
The most-starred other Audio projects in the LibreAI catalogue — the realistic shortlist if Qwen3-TTS is not the right fit.
利用 AI 大模型和自动化工作流，根据主题或关键词一键生成高清短视频。Generate HD short videos from a topic or keyword with an automated AI workflow.
Robust Speech Recognition via Large-Scale Weak Supervision
1 min voice data can also be used to train a good TTS model! (few shot voice cloning)
Clone a voice from 5 seconds of audio and generate arbitrary speech in real time
Port of OpenAI's Whisper model in C/C++
Free and open — grab the source, star the repo, and start creating.