Overview
What is Voicevox MCP Server?
Voicevox MCP Server lets you use VOICEVOX-compatible speech synthesis engines (AIVISpeech, VOICEVOX, COEIROINK) through the Model Context Protocol (MCP). It is designed for agent‑mode speech synthesis with Claude 3.7 in editors like Cursor.
How to use Voicevox MCP Server?
Clone the repository, run npm install, copy .env_example to .env, and set VOICEVOX_API_URL and VOICEVOX_SPEAKER_ID. On Windows, build with npm run build then start the server with npm start. Configure the editor’s mcp.json to point to the SSE endpoint. For Docker, run the container with stdio mode and the appropriate environment variables.
Key features of Voicevox MCP Server
- Supports VOICEVOX, AIVISpeech, and COEIROINK engines
- Adjustable speaker ID via environment variable
- Audio playback via VLC (Windows) or PulseAudio (Docker)
- Docker deployment with stdio transport
Use cases of Voicevox MCP Server
- Adding speech output to AI agent workflows in Cursor
- Using Claude 3.7 agent mode to generate spoken responses
- Integrating VOICEVOX-compatible TTS into MCP‑enabled applications
- Running TTS in a containerized environment
FAQ from Voicevox MCP Server
Which speech synthesis engines are supported?
It is compatible with VOICEVOX, AIVISpeech, and COEIROINK servers that expose a VOICEVOX‑compatible API.
How do I change the speaker voice?
Set the VOICEVOX_SPEAKER_ID environment variable. The default is 1 (Shikoku Metan). A list of available speaker IDs can be obtained from the engine’s /speakers endpoint.
What are the system requirements?
On Windows: Node.js 18+, a local VOICEVOX ENGINE instance, and VLC media player with its path set. On Docker/Linux: Docker, Docker Compose, WSL2, PulseAudio, and SDL2 libraries.
How does the server connect to the editor?
On Windows it uses SSE at http://localhost:10100/sse. On Docker it communicates over stdio.
How do I troubleshoot audio playback?
Ensure VLC is correctly installed and on the system path (Windows). For Docker, verify PulseAudio configuration and that the necessary packages (libsdl2-dev, pulseaudio-utils, pulseaudio) are available.
