This guide explains how to configure the Model Context Protocol (MCP) for use with LM Studio and helps you discover new MCP servers from the awesome-mcp repository.
LM Studio is a desktop application that allows you to run local LLMs with various model interfaces. The Model Context Protocol (MCP) enables LM Studio to connect to external tools, databases, and services through a standardized protocol.
- Operating System: Windows, macOS, or Linux
- LM Studio: Download from lmstudio.ai
- Node.js (Optional for advanced setups)
- Launch the LM Studio desktop application
- Go to Settings (βοΈ icon or menu)
- Navigate to the MCP Servers section
Click "Add Server" and provide:
- Name: A unique identifier for your server
- Command: The executable to run (e.g.,
npx,node) - Arguments: Additional flags or options
- Environment Variables (Optional): Set key-value pairs
Example configuration:
{
"name": "github-search",
"command": "npx",
"args": ["-y", "@modelcontextprotocol/server-filesystem"],
"env": {
"HOME": "${HOME}"
}
}The awesome-mcp repository is your go-to resource for discovering new and maintained Model Context Protocol servers! π
awesome-mcp is a curated collection of MCP servers that includes:
β
Official servers from the MCP specification maintainers
β
Community-built servers created by developers worldwide
β
Usage examples and documentation links
β
Maintenance status indicators to help you choose reliable projects
- Visit: https://github.com/terry-iverson/awesome-mcp
- Scroll through the categorized list of servers
- Check each repository's README for installation instructions
- Look for stars, forks, and recent commits as indicators of active maintenance
- Development Tools: Git, Docker, npm, etc.
- Cloud Services: AWS, GCP, Azure tools
- Databases: PostgreSQL, MongoDB, Redis connectors
- File System: File operations, cloud storage access
- Custom Servers: Community-specific implementations
Here's a complete example of adding an MCP server:
{
"id": "my-first-mcp-server",
"name": "GitHub Issues Helper",
"command": "npx",
"args": [
"-y",
"@modelcontextprotocol/server-git"
],
"env": {},
"transport": "stdio"
}- Use environment variables for sensitive data (API keys, tokens)
- Read each server's documentation before deploying
- Check security permissions - MCP servers can access your system
- Start with simple configurations to understand the protocol
- Join the community on GitHub or Discord for support
| Issue | Solution |
|---|---|
| Connection refused | Ensure the server executable exists and is in PATH |
| Authentication failed | Verify API keys/tokens are correctly set in environment variables |
| Command not found | Add npx -y for one-time installs or specify full paths |
- awesome-mcp Repository: https://github.com/terry-iverson/awesome-mcp
- LM Studio Documentation: https://lmstudio.ai/docs/
- MCP Specification: https://spec.modelcontextprotocol.io/
- LM Studio Discord: For community support
# Install a specific MCP server globally
npm install -g @modelcontextprotocol/server-name
# Run with npx (no global install needed)
npx -y @modelcontextprotocol/server-name
# Test your configuration
mcp-cli test-config --file path/to/config.jsonHappy configuring! π
Remember: awesome-mcp is always your best friend when looking for new MCP servers!