Awesome MCP Clients

Visit the repository to explore the full list of MCP clients with descriptions, license types, platform support, and links.

# Main repository:
# https://github.com/punkpeye/awesome-mcp-clients

# Live searchable directory:
# https://glama.ai/mcp/clients

The directory groups clients into categories: desktop applications, web-based platforms, CLI tools, IDE integrations, and bots. Identify which category fits your workflow — e.g., a VS Code user should look at IDE integrations like Cline or Continue.

Follow the installation instructions for your selected client. For example, Claude Desktop is installed from claude.ai/desktop, Cline is installed as a VS Code extension, and CLI-based clients can be installed via npm.

# Install Cline for VS Code:
code --install-extension saoudrizwan.claude-dev

# Install an npm-based CLI client:
npm install -g @anthropic-ai/claude-code

Most MCP clients use a JSON configuration file to register MCP servers. The configuration format is consistent across clients — a mcpServers object with named server entries specifying the command, args, and env.

{
  "mcpServers": {
    "filesystem": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-filesystem", "/path/to/files"]
    },
    "fetch": {
      "command": "uvx",
      "args": ["mcp-server-fetch"]
    }
  }
}

If you find an MCP client that is not listed, submit a PR to the repository following the existing format. The glama.ai directory also accepts submissions through its web interface.