ElevenLabs Voice To Voice Agent
БесплатноНе проверенAn MCP server that enables voice-to-voice AI conversations using ElevenLabs for speech synthesis and recognition, with tools for voice management, text-to-speec
Описание
An MCP server that enables voice-to-voice AI conversations using ElevenLabs for speech synthesis and recognition, with tools for voice management, text-to-speech, and speech-to-text.
README
A full voice-to-voice AI agent built on the Model Context Protocol. Speak into your microphone, the agent thinks, and ElevenLabs speaks back — all orchestrated through MCP tools, prompts, and resources.
Works with Anthropic Claude or OpenAI GPT — no Claude Desktop required.
MCP Primitives Implemented
Tools (7)
| Tool | Description |
|---|---|
list_voices |
List all available ElevenLabs voices |
get_voice |
Get detailed info about a specific voice |
list_models |
List available TTS/STT models |
text_to_speech |
Convert text → MP3, save to disk |
speech_to_text |
Transcribe an audio file → text |
get_user_info |
Account and character usage info |
get_history |
Retrieve generation history |
Prompts (4) — prompt-centric MCP
| Prompt | Args | What it does |
|---|---|---|
voice_agent_persona |
name, personality, [voice_style] |
Creates a named agent persona; picks the best matching voice and speaks every reply aloud |
start_voice_session |
[topic], [language], [voice_id] |
Initialises a live voice loop — agent listens and responds in spoken audio |
find_voice_for_role |
role |
Lists voices, picks top 3 for the role, generates audio samples for each |
voice_showcase |
phrase, [count] |
Renders the same phrase across N diverse voices for side-by-side comparison |
Resources (2)
| URI | Description |
|---|---|
elevenlabs://voices |
Live voice list — IDs, categories, labels |
elevenlabs://models |
Available TTS/STT models and capabilities |
Prerequisites
- Node.js ≥ 18
- sox (for microphone recording in
/voicemode)brew install sox # macOS sudo apt install sox # Ubuntu/Debian - An ElevenLabs API key
- An LLM key: Anthropic or OpenAI
Setup
npm install
npm run build
Environment Variables
| Variable | Required | Description |
|---|---|---|
ELEVENLABS_API_KEY |
Yes | ElevenLabs API key |
LLM_PROVIDER |
No | "anthropic" (default) or "openai" |
ANTHROPIC_API_KEY |
If using Anthropic | Claude API key |
OPENAI_API_KEY |
If using OpenAI | GPT API key |
ELEVENLABS_OUTPUT_DIR |
No | Where MP3 files are saved (defaults to cwd) |
Export before running (the server reads from process.env directly — .env is not auto-loaded):
export ELEVENLABS_API_KEY="your-key"
export LLM_PROVIDER="anthropic"
export ANTHROPIC_API_KEY="your-key"
Usage
Start the client
npm run client
ElevenLabs Voice Agent — MCP Client (provider: anthropic)
Connected! 7 tools: list_voices, get_voice, list_models, text_to_speech, speech_to_text, get_user_info, get_history
Commands: /prompts /prompt <name> [args] /voice exit
Text chat
You: List all available voices
[tool: list_voices]
Assistant: Here are the available voices...
You: Convert "Welcome to my demo" to speech using a deep male voice
[tool: list_voices]
[tool: text_to_speech]
Assistant: Audio saved to: ./output_1234.mp3
List and invoke prompts
You: /prompts
Available prompts:
voice_agent_persona — Create a voice AI agent with a specific persona...
args: name, personality, [voice_style]
start_voice_session — Initialize a live voice conversation...
args: [topic], [language], [voice_id]
find_voice_for_role — Analyze and recommend voices for a use case...
args: role
voice_showcase — Render a phrase in multiple voices...
args: phrase, [count]
You: /prompt voice_agent_persona name="Aria" personality="warm and professional"
[tool: list_voices]
[tool: text_to_speech]
Assistant: Hi, I'm Aria! How can I help you today?
[Audio plays automatically]
Voice conversation mode
You: /voice
[Voice Mode] The agent will listen and speak back.
[Voice Mode] Type 'exit' at any prompt to return to text mode.
Press ENTER to speak (or type 'exit'):
Recording... Press ENTER to stop:
Transcribing...
You: "What voices does ElevenLabs have?"
[tool: list_voices]
[tool: text_to_speech]
Assistant: ElevenLabs has over 30 voices...
[Audio plays automatically]
Press ENTER to speak (or type 'exit'): exit
[Returned to text mode]
ElevenLabs API Key Permissions
When creating your API key, enable:
| Permission | Setting |
|---|---|
| Text to Speech | Access |
| Speech to Text | Access |
| Voices | Read |
| Models | Access |
| History | Read |
| User | Read |
Everything else can stay at No Access.
With Claude Desktop
Add to claude_desktop_config.json:
{
"mcpServers": {
"elevenlabs": {
"command": "node",
"args": ["/absolute/path/to/build/index.js"],
"env": { "ELEVENLABS_API_KEY": "your-key" }
}
}
}
With Claude Code
claude mcp add elevenlabs node /absolute/path/to/build/index.js
Architecture
src/
elevenlabs-client.ts — ElevenLabs REST API wrapper (TTS + STT + voices + models)
index.ts — MCP server: 7 tools, 4 prompts, 2 resources
client.ts — CLI client: text REPL + /voice loop + /prompt commands
You (mic/keyboard)
│
▼
client.ts ──────────────────────────────► LLM API (Claude / GPT)
│ MCP stdio transport │
▼ │ tool calls
index.ts (MCP Server) ◄────────────────────────┘
│
▼
elevenlabs-client.ts ──► ElevenLabs REST API
(TTS + STT + voices + models)
Voice loop flow:
- Mic → sox → WAV file
speech_to_texttool → transcript text- LLM (with
VOICE_SYSTEMprompt) → text response +text_to_speechcall text_to_speechtool → MP3 fileafplay/mpg123→ audio plays from speakers- Loop back to step 1
License
MIT
from github.com/Begwen/Agentic-Voice-AI-System---MCP-Native-
Установка ElevenLabs Voice To Voice Agent
У этого сервера нет опубликованного пакета — он собирается из исходников. Открой репозиторий и следуй инструкции в README.
▸ github.com/Begwen/Agentic-Voice-AI-System---MCP-Native-FAQ
ElevenLabs Voice To Voice Agent MCP бесплатный?
Да, ElevenLabs Voice To Voice Agent MCP бесплатный — установка в пару кликов через Unyly без оплаты.
Нужен ли API-ключ для ElevenLabs Voice To Voice Agent?
Нет, ElevenLabs Voice To Voice Agent работает без API-ключей и переменных окружения.
ElevenLabs Voice To Voice Agent — hosted или self-hosted?
Self-hosted: сервер запускается локально на твоей машине командой из раздела установки.
Как установить ElevenLabs Voice To Voice Agent в Claude Desktop, Claude Code или Cursor?
Открой ElevenLabs Voice To Voice Agent на unyly.org, выбери вкладку своего клиента (Claude Desktop, Claude Code, Cursor) и нажми Install — конфиг сгенерируется автоматически, без правки JSON.
Похожие MCP
Opencode Omniroute Plugin
OpenCode plugin for the OmniRoute AI Gateway. Drives dynamic model discovery, /connect auth flow, and multi-instance OmniRoute providers via the official @openc
автор: GitHub ActionsRoblox Studio
Enables AI coding tools to control Roblox Studio for workspace exploration, instance manipulation, and script management. It provides tools for playtesting, sce
автор: paralovAntigravity Drawio
MCP server that enables AI assistants to create, parse, render, and validate Draw.io diagrams programmatically.
автор: S-SUJAN-SFetch
Web content fetching and conversion for efficient LLM usage.
AWS KB Retrieval
Retrieval from AWS Knowledge Base using Bedrock Agent Runtime.
автор: modelcontextprotocolSpring AI MCP Server
Provides auto-configuration for setting up an MCP server in Spring Boot applications.
llm-analysis-assistant
A very streamlined mcp client that supports calling and monitoring stdio/sse/streamableHttp, and can also view request responses through the /logs page. It also
автор: xuzexin-hzMCP-Agent
A simple, composable framework to build agents using Model Context Protocol by [LastMile AI](https://www.lastmileai.dev)
автор: lastmile-aiSpring AI MCP Client
Provides auto-configuration for MCP client functionality in Spring Boot applications.
mcp.natoma.ai
A Hosted MCP Platform to discover, install, manage and deploy MCP servers by [Natoma Labs](https://www.natoma.ai)
Compare ElevenLabs Voice To Voice Agent with
Не уверен что выбрать?
Найди свой стек за 60 секунд
Автор?
Embed-бейдж для README
Похожее
Все в категории ai
