elevenlabs/elevenlabs
MCP ServerElevenLabs MCP Server - text-to-speech, voice cloning
check: npx https://github.com/elevenlabs/elevenlabs-mcp
ElevenLabs text-to-speech MCP server. Generate speech, clone voices, transcribe audio, design voices, and create soundscapes.
MCP Server
- Package:
elevenlabs-mcp (Python — use uvx) - Transport: stdio
- Command:
uvx elevenlabs-mcp
Authentication
API key
| Variable | Required | Description |
|---|---|---|
ELEVENLABS_API_KEY | Yes | API key from elevenlabs.io/app/settings/api-keys |
ELEVENLABS_MCP_BASE_PATH | No | Default file location (~/Desktop) |
Available Tools
| Tool | Description |
|---|---|
text_to_speech | Generate speech from text |
voice_clone | Clone a voice |
transcribe_audio | Transcribe audio |
voice_design | Design new voice |
audio_isolation | Isolate audio |
soundscape | Create soundscape |
Configuration Example
[[mcp.servers]]
name = "elevenlabs"
type = "stdio"
command = "npx"
args = ["-y", "elevenlabs-mcp (Python — use uvx)"]
timeout_seconds = 60
env = { ELEVENLABS_API_KEY = "your-value", ELEVENLABS_MCP_BASE_PATH = "your-value" }
tools = []Notes: Python-based — use uvx, not npx. Free tier: 10,000 monthly credits.