home / mcp / elevenlabs mcp server

ElevenLabs MCP Server

Provides programmatic access to ElevenLabs TTS and audio APIs via MCP clients for speech, voice cloning, and transcription.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "nguyendinhsinh361-elevenlabs-mcp": {
      "command": "uvx",
      "args": [
        "elevenlabs-mcp"
      ],
      "env": {
        "ELEVENLABS_API_KEY": "<insert-your-api-key-here>"
      }
    }
  }
}

The ElevenLabs MCP Server lets MCP clients like Claude Desktop and other tools talk to ElevenLabs’ text-to-speech and audio processing APIs. It enables you to generate speech, clone voices, transcribe audio, and more by exposing a programmable interface that you can connect to from your preferred MCP client.

How to use

Set up an MCP-enabled workflow by first obtaining an API key from ElevenLabs. Then configure your MCP client to point at the ElevenLabs MCP server you run locally or remotely.

For Claude Desktop, you provide a configuration that launches the MCP server as a background process. The server runs under uvx and uses your ElevenLabs API key to access the services.

For other MCP clients, you typically generate a client-ready config by running a small helper command and then paste the resulting configuration into your client’s settings. This config tells the client how to start the MCP server and what API key to use.

How to install

Prerequisites you need before installation: a Python environment and the uvx tool (Python package manager) to run MCP servers.

Install uvx using the recommended script, then follow with the server command to start the ElevenLabs MCP server.

If you are using Claude Desktop, you will place a small JSON snippet into Claude’s MCP configuration area to point to the ElevenLabs MCP server.

If you are using other MCP clients, install the ElevenLabs MCP package for your client and generate the configuration with a print option to obtain the ready-to-paste settings.

Additional notes and troubleshooting

Environment variable for the API key is ELEVENLABS_API_KEY. Provide your key in the configuration where shown, and keep it secure.

If you encounter path or executable errors, verify the absolute path to the MCP launcher (for example, the uvx binary) and update the configuration accordingly.

You can explicitly set a base path for MCP I/O using ELEVENLABS_MCP_BASE_PATH to control where input and output files are read from and written to.

Available tools

generateSpeech

Generate speech audio from text using ElevenLabs TTS capabilities via MCP.

cloneVoice

Create a cloned voice profile by applying voice characteristics to an existing model.

transcribeAudio

Transcribe spoken audio to text using ElevenLabs transcription features.

audioProcessing

Apply audio processing effects or transformations as supported by ElevenLabs APIs.