home / mcp / google search mcp server

Google Search MCP Server

Provides Google Custom Search functionality via MCP for querying search results programmatically.

Installation

Add the following to your MCP client configuration file.

Configuration

{
  "mcpServers": {
    "renoscriptdev-mcp-google-scholar-dev": {
      "command": "mcp",
      "args": [
        "run",
        "google_search_mcp_server.py"
      ],
      "env": {
        "GOOGLE_CSE_ID": "YOUR_CSE_ID",
        "GOOGLE_API_KEY": "YOUR_API_KEY"
      }
    }
  }
}

A dedicated MCP server that exposes Google Custom Search functionality to MCP clients. It enables you to perform Google-powered searches through a consistent MCP interface, making it easy to integrate search results into your assistant workflows or applications without embedding direct Google API calls in your client code.

How to use

You interact with this server through an MCP client. Start the server and then issue search requests via the client’s standard MCP workflow. The server handles Google Custom Search API calls and returns structured results that you can present to users or feed into your application logic.

How to install

Prerequisites you need on your machine before installation:

- Python 3.x installed on your system

- pip for Python package installation

- Optional: Node.js if you plan to use Smithery for automatic installation

Install via Smithery

npx -y @smithery/cli install @gradusnikov/google-search-mcp-server --client claude

Manual installation

Clone the repository and install dependencies. The following steps guide you through a local setup intended for Python-based MCP servers.

git clone https://github.com/gradusnikov/google-search-mpc-server.git
cd google-search-mpc-server
```

```
pip install fastmcp google-api-python-client python-dotenv

Run the server

mcp run google_search_mcp_server.py

Integrate with Claude Desktop (optional)

If you are using Claude Desktop, add the server as an MCP entry in your CLAUDE_DIRECTORY/claude_desktop_config.json. For example, on Windows Subsystem for Linux (WSL) it may look like this:

"google-search": {
            "command": "wsl.exe",
            "args": [
                "bash",
                "-c",
                "source /home/[user]/anaconda3/etc/profile.d/conda.sh && conda activate mcp && /home/[user]/anaconda3/bin/mcp run /home/[user]/google-search-mpc-server/google_search_mcp_server.py"
            ]
        },