π¦
Voice MCP
An MCP (Model Context Protocol) server for AI voice synthesis with an inline audio player. Give your AI assistant a custom cloned voice!
0 installs
Trust: 34 β Low
Content
Ask AI about Voice MCP
Powered by Claude Β· Grounded in docs
I know everything about Voice MCP. Ask me about installation, configuration, usage, or troubleshooting.
0/500
Loading tools...
Reviews
Documentation
ποΈ voice-mcp
An MCP (Model Context Protocol) server for AI voice synthesis with an inline audio player. Give your AI assistant a custom cloned voice!
Features
- π€ Custom Voice Cloning β Use MiniMax TTS API with your own cloned voice
- π΅ Inline Audio Player β Beautiful WeChat-style player with waveform visualization
- π Transcript Toggle β Show/hide the spoken text
- π Dark Mode Support β Automatic theme adaptation
- β‘ Cloudflare Workers β Fast, serverless deployment
Demo
When you call the speak tool, you get:
- A sleek audio player with play/pause button
- Animated waveform that follows playback progress
- Duration display
- Expandable transcript
Quick Start
1. Clone the repository
git clone https://github.com/garan0613/voice-mcp.git
cd voice-mcp
2. Install dependencies
npm install
3. Configure MiniMax API
You'll need a MiniMax account with voice cloning enabled.
Add your secrets to Cloudflare:
npx wrangler secret put MINIMAX_API_KEY
npx wrangler secret put MINIMAX_GROUP_ID
npx wrangler secret put VOICE_ID
npx wrangler secret put BOT_NAME # Optional, defaults to "AI"
4. Deploy
npx wrangler deploy
5. Connect to Claude.ai
- Go to Settings β Connectors β Add Connector
- Enter your Worker URL:
https://your-worker.workers.dev/mcp - Done! The
speaktool is now available.
Configuration
| Variable | Required | Description |
|---|---|---|
MINIMAX_API_KEY | β | Your MiniMax API key |
MINIMAX_GROUP_ID | β | Your MiniMax group ID |
VOICE_ID | β | The cloned voice ID |
BOT_NAME | β | Display name (default: "AI") |
API Endpoints
| Endpoint | Description |
|---|---|
GET /mcp | MCP server (SSE protocol) |
GET /speak?text=Hello | Direct audio file |
GET /status | Health check |
How to Clone a Voice
- Go to MiniMax Console
- Navigate to Voice Cloning
- Upload 10-30 seconds of clear audio
- Wait for processing (usually a few minutes)
- Copy the Voice ID
Custom Deployment
Using a Custom Domain
- Add your domain to Cloudflare
- Create a DNS record pointing to your Worker
- Update
wrangler.jsonc:
{
"routes": [
{ "pattern": "voice.yourdomain.com/*", "zone_name": "yourdomain.com" }
]
}
Self-Hosting (Node.js)
The core MCP logic can be adapted for other platforms. You'll need to:
- Replace
createMcpHandlerwith a standard HTTP/SSE handler - Use
@modelcontextprotocol/sdkdirectly - Handle the SSE transport yourself
Tech Stack
- Cloudflare Workers β Serverless runtime
- MCP SDK β Model Context Protocol
- MiniMax TTS β Voice synthesis
- ext-apps β Inline UI rendering
License
MIT Β© 2026
Credits
Inspired by the need to give AI assistants a voice. Built with β€οΈ
