What it does
The official ElevenLabs MCP server that integrates text-to-speech, voice cloning, audio transcription, and voice conversion into MCP clients. It exposes ElevenLabs' audio generation and processing APIs—speech synthesis, voice cloning and conversion, audio transcription, and voice variation generation—to Claude, Cursor, Windsurf, and other MCP-compatible tools, enabling users to manipulate audio without leaving their development environment.
Who it's for
Voice engineers, content creators, and developers building audio-enhanced applications who want Claude to programmatically generate speech, clone or convert voices, or transcribe audio without switching between tools or manually uploading files.
Common use cases
- Generate dialogue or narration for video games, educational content, or podcasts using Claude-selected voices
- Clone a voice from a sample recording and then generate speech in that voice
- Transcribe audio recordings and regenerate the speech with different voice characteristics
- Create voice variations for characters in interactive fiction or game scenarios
- Convert one speaker's voice to another for accessibility or creative audio effects
Setup pitfalls
- Requires a valid ElevenLabs API key (free tier: 10,000 credits/month; paid plans unlock higher limits)
- Must install
uv(Python package manager) before installing the server viauvx - File output modes (
ELEVENLABS_MCP_OUTPUT_MODE) default to writing to disk at~/Desktop; specifyresourcesmode for containerized or serverless deployments - Last commit was 141 days ago; verify compatibility with your Claude version before relying on it in production