yovoice
Give your words a voice. Compose your ideas into songs.
English · 简体中文
Website · CLI & Agent Skill · API / MCP Deployment · Development
yovoice is an open-source, local-first audio and music creation studio for macOS and Windows.
Turn text into expressive speech, natural cloned voices, complete vocal songs, and multi-track scores right on your machine. Featuring zero-shot voice cloning, emotional performance controls, AI music generation with lyrics, and General MIDI score arranging—all running 100% locally with zero cloud API calls, zero per-character fees, and zero Python environment headaches.
Why yovoice?
- 🔒 100% Local Privacy: Runs entirely on your local GPU/CPU. Your manuscripts, reference voices, and creations never leave your computer.
- 💸 Zero Recurring Fees: Say goodbye to per-character/per-minute meters and expensive cloud voice/music subscriptions.
- ⚡ Native C++ Performance (No Python Hell): Powered by audio.cpp v0.9.1. Ships with bundled Windows CUDA 12.4/13.3 runtimes and native macOS Metal support. Double-click and create—no Conda or PyTorch conflicts.
- 🎛️ Creator Workstation: Multi-take versioning for side-by-side comparison, waveform trimming, timeline scrolling, and clean stems/MIDI export.
- 🤖 MCP & AI Agent Ecosystem: Connect DeepSeek, Qwen, Claude, OpenAI, or local Ollama to write scores and lyrics in seconds. Includes a native Model Context Protocol (MCP) server for Cursor and Claude Code.
Features
- Voice and expression — use a reference voice, match a reference performance, adjust emotions, or describe the delivery in words.
- A complete audio workflow — import or record reference audio, trim clips, preview speech, and export your work.
- Voice design and cloning — VoxCPM2 offers text-guided voice design, controllable cloning, and transcript-assisted cloning with automatic multilingual handling, and 48 kHz output.
- Music generation — describe a style, write lyrics with section markers, and generate full songs or instrumentals with ACE-Step 1.5; every take is kept as a version to compare.
- Score — write each part by bar and beat, import MIDI or let an Agent write the score, then render it with a General MIDI sound font and a role-based mix; results are exactly repeatable and export to MIDI and stems. See Score.
- Optional AI service — connect DeepSeek, Qwen, Kimi, OpenAI, Anthropic, Ollama or any compatible API to let AI write arrangements and lyrics; keys stay on this computer. See AI service.
- Local models — run IndexTTS 2.0 / 2.5, VoxCPM2, OmniVoice, Qwen3-TTS, Breeze TTS 2 and Kokoro through audio.cpp, with resumable model downloads and GGUF import.
- Hardware acceleration — Metal on Apple Silicon; CPU, NVIDIA CUDA, and experimental Vulkan on Windows.
- Agent Skill & MCP — ask your AI agent to set up local speech generation and create voiceovers from text and reference audio.
Hardware & VRAM Quick Guide
| Use Case | Recommended Hardware | Recommended Models |
|---|---|---|
| 🟢 Lightweight & Blazing Fast | Any CPU / 4 GB+ RAM | Kokoro-82M (only 189 MB Q8, millisecond latency, runs great on CPU) |
| 🟡 High-Fidelity Voice Cloning | 4 GB – 6 GB VRAM / 8 GB+ Apple Silicon | VoxCPM2 (2B, 48 kHz), IndexTTS 2.0/2.5, Qwen3-TTS |
| 🔴 Full Song & Music Generation | 12 GB+ VRAM / 16 GB+ Apple Silicon | ACE-Step 1.5 Turbo (BF16, 10 GB), YuE2 3B |
Supported Models
| Model | Parameters | Model file size by precision | Core capabilities |
|---|---|---|---|
| IndexTTS 2.0 | — | Q8 · 3.63 GB F16 · 4.65 GB ORIG · 8.08 GB |
Chinese/English voice cloning, emotion control, reference performance |
| IndexTTS 2.5 | — | Q8 · 3.50 GB F16 · 4.55 GB ORIG · 7.89 GB |
Multilingual voice cloning, emotion control, pronunciation editing |
| VoxCPM2 | 2B | Q8 · 2.96 GB BF16 · 4.77 GB ORIG · 4.96 GB |
Text-guided voice design, voice cloning, transcript-assisted cloning |
| OmniVoice | 0.6B | Q8 · 1.35 GB BF16 · 1.64 GB F16 · 1.64 GB |
Attribute-based voice design, voice cloning, non-verbal sound tags |
| Qwen3-TTS Base | 0.6B | Q8 · 1.99 GB BF16 · 2.52 GB |
Reference voice cloning, optional transcript guidance, multilingual speech |
| Qwen3-TTS Base | 1.7B | Q8 · 2.70 GB BF16 · 4.20 GB ORIG · 4.54 GB |
Reference voice cloning, optional transcript guidance, multilingual speech |
| Qwen3-TTS CustomVoice | 1.7B | Q8 · 2.82 GB BF16 · 4.18 GB |
9 built-in voices, text-guided style and emotion |
| Qwen3-TTS VoiceDesign | 1.7B | Q8 · 2.82 GB BF16 · 4.18 GB |
Voice design from natural-language descriptions, no reference audio required |
| Breeze TTS 2 | ~3B | Q4 · 3.79 GB Q8 · 5.08 GB BF16 · 7.34 GB |
Chinese and English voice design, cloning and voice direction; GPU recommended, requires the v0.9.1 engine |
| Kokoro-82M 1.0 Official | 82M | Q8 · 189.55 MB BF16 · 211.95 MB |
49 built-in voices, multilingual speech excluding Japanese |
| Kokoro-82M 1.0 | 82M | Q8 · 932.66 MB | 54 built-in voices, full multilingual resources including Japanese; import only |
| Kokoro-82M 1.1-zh | 82M | Q8 · 255.32 MB | 100 Chinese and 3 English voices; experimental, import only |
| ACE-Step 1.5 Turbo | — | BF16 · 10.09 GB | Songs and instrumentals from a style prompt and lyrics, 50+ vocal languages; requires engine v0.9.1 |
All models are available in the App and CLI. See model selection for input requirements and examples. OmniVoice (CC-BY-NC) and Breeze TTS 2 (BreezeBlue Research and Non-Commercial License) weights are restricted to non-commercial use.
Installation
Choose the package for your platform from the repository’s Releases tab or visit the official website.
| Platform | Package | Install |
|---|---|---|
| macOS 14+ · Apple Silicon | .dmg |
Open the disk image and drag yovoice to Applications |
| Windows 10/11 · x64 | -setup.exe |
Run the installer, then open yovoice from the Start menu |
| Windows 10/11 · x64 · NVIDIA GPU | -cuda12.4-setup.exe |
Bundles the CUDA 12.4 runtime; works with older GPUs and drivers and uses the GPU from first launch |
| Windows 10/11 · x64 · NVIDIA GPU | -cuda13.3-setup.exe |
Bundles the CUDA 13.3 runtime; needs a newer driver, smaller download |
The Windows installer downloads and installs WebView2 Runtime if needed. Models are downloaded inside the app; uninstalling preserves user data in ~/.yovoice.
Check for updates from the app menu on macOS or Windows. Updates are also checked and downloaded in the background; restart to install when ready. Windows updates download only the app (-update.exe) and keep the installed runtime; when a release recommends a newer runtime you can upgrade it in Settings → Inference engine.
Quick Start
- Set up a model. Download a model in Settings (e.g. Kokoro-82M for fast narration, or VoxCPM2 for cloning).
- Create a project. Choose Story dubbing, Speech generation, Music, or Score.
- Add content. Enter text, lyrics, or notes, then pick a character and adjust style.
- Generate audio. Select Generate, then preview, compare takes, edit, and export.
CLI & Agent Skill
Give your agent the yovoice Skill link and ask it to install the Skill and set up the local CLI, engine, and model:
Install this Skill and set up yovoice for local speech generation: https://github.com/leemysw/yovoice/tree/main/skills/yovoice
Then describe what you want:
Read narration.txt using voice.wav as the reference voice, with a calm delivery, and save it as narration.wav.
The agent runs the standalone CLI without opening the desktop app. See the CLI guide for details.
Run yovoice serve to provide an authenticated HTTP API and MCP inference service for other machines. See the API / MCP deployment guide for deployment and client examples.
Development
make install make app-run
See the development guide for prerequisites. It also covers browser preview, tests, and project structure.
Contributing
Bug reports, feature suggestions, and pull requests are welcome. Include your platform, model, and steps to reproduce when reporting a problem. Run make check before submitting code changes.
Acknowledgements
- audio.cpp by ShugoAI — the local audio inference engine.
- VoxCPM — voice design and cloning models, licensed under Apache-2.0.
- IndexTTS — the speech synthesis models behind yovoice.
- OmniVoice — voice design and cloning models.
- Qwen3-TTS — voice cloning, built-in voices, and text-guided voice design.
- ACE-Step — music generation from style prompts and lyrics.
- go-meltysynth, MuseScore General, GeneralUser GS and SGM-V2.01 by David Shan — the SoundFont synthesizer and sound fonts behind score rendering.
- Kokoro — lightweight speech synthesis with built-in voices: Kokoro-82M and Kokoro-82M-v1.1-zh.
Star History
License
Apache-2.0. Dependencies retain their original licenses; IndexTTS models are covered separately by their model license.