Kagi Search MCP Server

Overview

Kagi MCP server provides access to Kagi's search engine and content summarization tools through the Model Context Protocol. This integration allows AI assistants to perform web searches and generate summaries of various content types including articles, videos, and websites.

Prerequisites

Before using the Kagi MCP server, you'll need:

• A Kagi API key (especially for search functionality, which is currently in closed beta)
• For search access, contact support@kagi.com to request an invite
• The uv package manager installed on your system

Installation

Installing uv

First, install the uv package manager:

MacOS/Linux:

curl -LsSf https://astral.sh/uv/install.sh | sh

Windows:

powershell -ExecutionPolicy ByPass -c "irm https://astral.sh/uv/install.ps1 | iex"

Method 1: Installing via Smithery

The simplest way to install Kagi MCP is through Smithery:

npx -y @smithery/cli install kagimcp --client claude

Method 2: Installing via uvx

You can install and run the Kagi MCP server directly using uvx:

uvx kagimcp

Configuration

Claude Desktop Configuration

Add the following to your Claude Desktop configuration file:

1. Access your config file through: Hamburger Menu → File → Settings → Developer → Edit Config
2. Add the Kagi MCP server configuration:

{
  "mcpServers": {
    "kagi": {
      "command": "uvx",
      "args": ["kagimcp"],
      "env": {
        "KAGI_API_KEY": "YOUR_API_KEY_HERE",
        "KAGI_SUMMARIZER_ENGINE": "YOUR_ENGINE_CHOICE_HERE"
      }
    }
  }
}

Advanced Configuration Options

• Logging Level: Adjust the logging level using the FASTMCP_LOG_LEVEL environment variable (e.g., "ERROR", "INFO", etc.)
• Summarizer Engine: Customize the summarization engine using the KAGI_SUMMARIZER_ENGINE environment variable
  • Default: "cecil"
  • Other options include "daphne" and others (see Kagi documentation for details)

Debugging

To debug the Kagi MCP server, run:

npx @modelcontextprotocol/inspector uvx kagimcp

This will launch the MCP Inspector interface at http://localhost:5173. You may need to add your Kagi API key in the environment variables section under KAGI_API_KEY.

Usage Examples

Once configured, you can use the Kagi MCP server by asking questions that require search or summarization:

• For search: "Who was Time's 2024 person of the year?"
• For summarization: "Summarize this video: https://www.youtube.com/watch?v=jNQXAC9IVRw"

Local Development Setup

For developers who want to work with the source code:

1. Clone the repository:

   git clone https://github.com/kagisearch/kagimcp.git
   

2. Install dependencies:

   cd kagimcp
   uv venv
   source .venv/bin/activate  # MacOS/Linux
   # OR
   .venv/Scripts/activate  # Windows
   uv sync
   

3. Configure Claude Desktop for local development using either:

   • MCP CLI SDK:

     pip install mcp[cli]
     mcp install /ABSOLUTE/PATH/TO/PARENT/FOLDER/kagimcp/src/kagimcp/server.py -v "KAGI_API_KEY=API_KEY_HERE"
     

   • Manual configuration (see Configuration section above, but use the local path)