WhatsApp Cloned Voice Messages

Integrates WhatsApp and Minimax to send personalized voice messages using cloned voices.

Complete Guide: Claude Desktop MCP Setup for WhatsApp Voice Messages with Cloned Voices

This comprehensive guide documents the complete process of setting up Claude Desktop with MCP (Model Context Protocol) servers for WhatsApp and Minimax integration. The goal is to enable sending personalized voice messages using cloned voices directly through WhatsApp from Claude Desktop.

- ✅ Configure Claude Desktop with WhatsApp MCP server
- ✅ Integrate Minimax MCP server for AI voice generation
- ✅ Clone your voice and use it for message generation
- ✅ Send custom audio messages directly to WhatsApp contacts
- ✅ Handle common setup issues and troubleshooting

- macOS or Linux system
- Claude Desktop installed
- Python 3.8+ installed
- Git installed
- WhatsApp account
- Minimax account with API access

- WhatsApp MCP Server:https://github.com/lharries/whatsapp-mcp/tree/main
- Minimax MCP Server:https://github.com/MiniMax-AI/MiniMax-MCP
- Video example:https://www.youtube.com/watch?v=1xwAl284fJ8

# Install uv curl -LsSf https://astral.sh/uv/install.sh | sh # Add to PATH (add to your ~/.bashrc or ~/.zshrc) export PATH="$HOME/.local/bin:$PATH" # Reload shell or run: source ~/.bashrc # or ~/.zshrc

# Clone the repository cd ~/ mkdir -p github_com_lharries_whatsapp-mcp cd github_com_lharries_whatsapp-mcp git clone https://github.com/lharries/whatsapp-mcp.git whatsapp-mcp-server cd whatsapp-mcp-server

# Install Python dependencies using uv uv sync

- QR Code not appearing: Check firewall settings and ensure port access
- Authentication fails: Try clearing browser cache and regenerating QR code
- Session expires: You may need to re-authenticate periodically
- Sign up atMinimax AI platform
- Navigate to API settings
- Generate your API key
- Save the key securely (you'll need it for configuration)

# Install using uvx uvx install minimax-mcp

# Clone repository method git clone https://github.com/MiniMax-AI/MiniMax-MCP.git cd MiniMax-MCP uv sync

{ "mcpServers": { "whatsapp": { "command": "/Users/[YOUR_USERNAME]/.local/bin/uv", "args": [ "--directory", "/Users/[YOUR_USERNAME]/github_com_lharries_whatsapp-mcp/whatsapp-mcp-server", "run", "main.py" ] }, "MiniMax": { "command": "/Users/[YOUR_USERNAME]/.local/bin/uvx", "args": [ "minimax-mcp", "-y" ], "env": { "MINIMAX_API_KEY": "YOUR_MINIMAX_API_KEY_HERE", "MINIMAX_MCP_BASE_PATH": "/tmp", "MINIMAX_API_HOST": "https://api.minimaxi.chat", "MINIMAX_API_RESOURCE_MODE": "local" } } } }

Important: Replace[YOUR_USERNAME]with your actual username andYOUR_MINIMAX_API_KEY_HEREwith your actual Minimax API key.

- Ensure all paths use absolute paths
- Verifyuvanduvxare installed in~/.local/bin/
- Check that directory structure matches exactly

- Validate JSON syntax using online validators
- Ensure all quotes are properly escaped
- Check for trailing commas

- Verify API key is correct and active
- Ensure base path/tmpexists and is writable
- Check API host URL is accessible

Step 5: Voice Cloning Setup (Optional but Recommended)

Once Claude Desktop is running with MCP servers:

Can you list all available voices in Minimax?

Can you clone my voice using this audio file? [upload your voice sample] Use voice ID: my_custom_voice_[your_name]

Generate a test audio message with my cloned voice saying "Hello, this is a test message"

- Completely close Claude Desktop
- Restart the application
- Wait for MCP servers to initialize
- Check that both servers are connected (you should see MCP indicators in Claude)

Can you search for contacts in my WhatsApp?

Can you generate an audio message saying "Hello, this is a test" using my cloned voice?

Can you send a voice message to [phone number] saying "Hello, this is a test message from Claude" using my cloned voice?

Here's the exact process we followed that resulted in success:

User: "Can you send an audio message with my cloned voice to +34959217658 with the message 'Hemos llegado a los 140.000 seguidores en instagram. Que chimba!! Gracias a la divinidad'"

Claude listed available voices and found the cloned voice: moss_audio_17d8fcf2-4abf-11f0-b862-46ba4da2d9df

Parameters used: - Text: "Hemos llegado a los 140.000 seguidores en instagram. Que chimba!! Gracias a la divinidad" - Voice ID: moss_audio_17d8fcf2-4abf-11f0-b862-46ba4da2d9df - Emotion: happy - Language: Spanish

Sent to: 34959217658 (without + prefix) Format: Audio message Result: Success

User requested change from "140.000 seguidores" to "140 mil seguidores" Regenerated audio with corrected text Sent updated message successfully

Symptoms: Claude Desktop shows no MCP servers connected
- Check configuration file syntax
- Verify all paths are absolute and correct
- Ensureuvanduvxare properly installed
- Restart Claude Desktop completely
- Check system logs for error messages

Symptoms: Cannot connect to WhatsApp, QR code issues
- Clear WhatsApp Web sessions in your phone
- Ensure good internet connection
- Try running WhatsApp MCP server manually first
- Check firewall/antivirus settings

Symptoms: Voice generation fails, API errors
- Verify API key is correct and active
- Check account credit balance
- Ensure API host URL is accessible
- Try different base path if permission issues

Symptoms: Cannot find cloned voice, generation fails
- Ensure voice was properly cloned in Minimax platform
- Check voice ID format
- Verify audio file format compatibility
- Try listing voices to confirm availability

Symptoms: Audio generates but WhatsApp sending fails
- Check phone number format (no + prefix)
- Ensure WhatsApp session is active
- Verify audio file format compatibility
- Check file permissions and paths
- Regular Backups: Save your Claude Desktop configuration
- API Key Security: Never share your Minimax API key
- Voice Quality: Use high-quality audio for voice cloning
- Testing: Always test with yourself first before sending to others
- Updates: Keep MCP servers updated regularly

cd ~/github_com_lharries_whatsapp-mcp/whatsapp-mcp-server uv run main.py

cat ~/Library/Application\ Support/Claude/claude_desktop_config.json | python -m json.tool

You'll know everything is working when:

- ✅ Claude Desktop shows both MCP servers connected
- ✅ You can search WhatsApp contacts
- ✅ You can list available voices including your cloned voice
- ✅ You can generate audio with your voice
- ✅ You can send audio messages to WhatsApp successfully

This setup enables powerful automation capabilities by combining Claude's intelligence with WhatsApp's reach and Minimax's voice cloning technology. The key to success is careful attention to configuration details and systematic troubleshooting of any issues that arise.

The complete integration allows for natural language commands to Claude that result in personalized voice messages being sent directly to WhatsApp contacts, opening up possibilities for automated customer service, personal assistants, and creative content distribution.

An MCP server integrating WhatsApp messaging and ElevenLabs AI voice capabilities into VS Code.

An AI voice toolkit with TTS, voice cloning, and video translation, now available as an MCP server for smarter agent integration.

A server for text-to-speech generation using the AivisSpeech engine.

A Node.js MCP server for the Japanese text-to-speech software Bouyomi-chan.

Generates text-to-speech audio with automatic playback using the Chatterbox TTS model.

Text-to-speech generation with conversation history features using the ElevenLabs API.

Voice MCP server: voice-cloning TTS/STT with speak, talk, voice inbox and feeds - remote MCP with OAuth at mcp.koe.live

A server that provides text-to-speech capabilities using the Kokoro TTS engine.

Model Context Protocol (MCP) server for Kokoro text-to-speech with female voice. 100% local, no Python required. Supports SSE and stdio transports.

Create AI-generated memes and convert them into stickers for Telegram and WhatsApp.