Serper (Google Search)

🚀 Serper MCP Server

A Model Context Protocol (MCP) server that provides Google Search via Serper. This server enables large language models (LLMs) to access Google search result information.

✨ Features

Available Tools

• google_search - Retrieve search results.
  • Required parameters:
    • q (string): The query to search for.
  • Optional parameters:
    • gl (string): The country where the search is conducted, e.g., us, uk, ca, au, etc.
    • location (string): The location of the search, e.g., San Francisco, CA, USA.
    • hl (string): The language of the search, e.g., en, es, fr, de, etc.
    • tbs (string): The time range of the search, e.g., d (day), w (week), m (month), y (year).
    • num (integer): The number of results to return, with a maximum of 100 (default: 10).
    • page (integer): The page number to return, with the first page being 1 (default: 1).

🚀 Quick Start

Using uv (Recommended)

1. Ensure that you have installed uv on your operating system.
2. Add the serper MCP server to your MCP client code configuration file or Claude settings (file claude_desktop_config.json):

   {
       "mcpServers": {
           "serper": {
               "command": "uvx",
               "args": ["serper-mcp-server"],
               "env": {
                   "SERPER_API_KEY": "<Your Serper API key>"
               }
           }
       }
   }
   

   uv will automatically download the MCP server from pypi.org and apply it to your MCP client.

Using pip in a Project

1. Add serper-mcp-server to the requirements.txt file of your MCP client code:

   serper-mcp-server
   

2. Install the dependencies.

   pip install -r requirements.txt
   

3. Add the client configuration:

   {
       "mcpServers": {
           "serper": {
               "command": "python3",
               "args": ["-m", "serper_mcp_server"],
               "env": {
                   "SERPER_API_KEY": "<Your Serper API key>"
               }
           }
       }
   }
   

Using pip Globally

1. Ensure that you have pip or pip3 installed on your system.

   pip install serper-mcp-server
   # or
   pip3 install serper-mcp-server
   

2. Add the serper MCP server to your MCP client code configuration file or Claude settings:

   {
       "mcpServers": {
           "serper": {
               "command": "python3",
               "args": ["serper-mcp-server"],
               "env": {
                   "SERPER_API_KEY": "<Your Serper API key>"
               }
           }
       }
   }
   

💻 Usage Examples

Debugging

You can use the MCP inspector to debug the server. For a uvx installation:

npx @modelcontextprotocol/inspector uvx serper-mcp-server

Or if you have installed the package in a specific directory or are developing:

cd path/to/your/directory
npm install -g @modelcontextprotocol/inspector
@modelcontextprotocol/inspector uvx serper-mcp-server

📚 Documentation

Important Note

⚠️ Important Note

Please ensure that you have obtained a valid Serper API key and correctly set the environment variables as per the documentation.