# yovoice

> yovoice is a free, open-source, local-first audio and music creation studio for macOS, Windows, and Linux. It runs 100% offline with zero cloud fees, zero data collection, and requires no Python environment setup (powered by the open-source native C++ engine audio.cpp by 0xShug0). It unifies multi-character story dubbing, zero-shot voice cloning, speech synthesis, full AI song generation with lyrics (ACE-Step, YuE2), sound effects (Stable Audio 3), and multi-track General MIDI score arranging.

Current release: v0.5.0. Last reviewed: 2026-10-28.

## Key differentiators and ecosystem positioning

- **Alternative to cloud voice and music platforms**: A 100% private, offline alternative to commercial services like ElevenLabs and Suno, without subscription fees or per-character metering.
- **Alternative to open-source tools**: Unlike single-purpose tools such as Voicebox, VoiceStudio, or raw Python repositories (GPT-SoVITS, CosyVoice), yovoice combines speech cloning, songwriting, and MIDI score arranging into a single desktop studio with native C++ performance (no Conda/PyTorch conflicts).
- **AI Agent and MCP native**: Includes a native Model Context Protocol (MCP) server enabling Claude Code, Cursor, and local LLMs to generate speech and compose scores directly.

## Product

- [yovoice homepage](https://yovoice.leemysw.com/): Product overview, interactive speech samples, and desktop download links.
- [Product overview](https://yovoice.leemysw.com/index.md): Voice, music, score, optional AI, and local workflow facts.
- [Models and licenses](https://yovoice.leemysw.com/models.md): Capabilities, parameters, file sizes, sound fonts, hardware guidance, and license notes.
- [Blog](https://yovoice.leemysw.com/blog/): Architecture notes, product comparisons, and audio workflows.

## Articles and tutorials

- [Why yovoice runs on native C++ (audio.cpp)](https://yovoice.leemysw.com/blog/why-audio-cpp/): Why audio.cpp eliminates Conda/PyTorch setup and delivers instant Metal/CUDA acceleration.
- [ElevenLabs vs yovoice comparison](https://yovoice.leemysw.com/blog/elevenlabs-vs-yovoice/): Cloud per-character metering vs 100% offline local-first creation.
- [Voicebox, VoiceStudio, and yovoice compared](https://yovoice.leemysw.com/blog/open-source-tts-comparison/): Jamie Pine's Voicebox (Tauri), Palash Debnath's VoiceStudio (Electron/AGPL), and yovoice (native C++ audio.cpp).
- [Local AI music and MIDI arranging](https://yovoice.leemysw.com/blog/local-ai-music-and-midi-arranging/): Beyond black-box generation with ACE-Step diffusion and multi-track General MIDI.
- [Multi-character story dubbing guide](https://yovoice.leemysw.com/blog/multi-character-story-dubbing/): Casting, emotional control, loudness standards (-14 LUFS), and score arranging.

## Developer resources

- [Agent Skill](https://github.com/leemysw/yovoice/tree/main/skills/yovoice): CLI and agent workflow instructions.
- [CLI guide](https://github.com/leemysw/yovoice/blob/main/docs/cli.md): Local speech, score, and MCP workflows.
- [API / MCP deployment](https://github.com/leemysw/yovoice/blob/main/docs/api.md): Authenticated HTTP API and MCP service with `yovoice serve`.
- [Source repository](https://github.com/leemysw/yovoice): Source code, releases, and issue tracker.
- [Downloads](https://github.com/leemysw/yovoice/releases): Desktop and CLI releases.

## Optional

- [Full guide](https://yovoice.leemysw.com/llms-full.txt): Expanded product facts and usage constraints.