How to Integrate Brave Search MCP Server with Your AI Agent

How to Integrate Brave Search MCP Server with Your AI Agent

This tutorial walks you through setting up the Brave Search MCP Server and connecting it to MCP-compatible clients like Claude Desktop, Cursor, or any agent framework that supports the Model Context Protocol. By the end, your AI agent will be able to search the web in real-time and retrieve live information.

Prerequisites

• Node.js 18+ or Docker (for running the server)
• A Brave Search API key (free tier: 2,000 queries/month)
• An MCP-compatible client (Claude Desktop, Cursor IDE, or any MCP host)

Step 1: Get a Brave Search API Key

Sign up at the Brave Search API website and choose a plan (the free tier offers 2,000 queries per month — plenty for most use cases). Once registered, generate your API key from the developer dashboard. Keep this key handy — you'll need it in the next step.

Step 2: Run the Server

Option A: Using Docker (Recommended)

docker pull mcp/brave-search:latest
docker run -e BRAVE_API_KEY=your-api-key-here mcp/brave-search:latest

Option B: Using npx

env BRAVE_API_KEY=your-api-key-here npx @modelcontextprotocol/server-brave-search

Step 3: Configure Your MCP Client

Add the server to your Claude Desktop config at ~/Library/Application Support/Claude/claude_desktop_config.json (macOS):

{
  "mcpServers": {
    "brave-search": {
      "command": "docker",
      "args": ["run", "-i", "--rm", "-e", "BRAVE_API_KEY", "mcp/brave-search"],
      "env": {
        "BRAVE_API_KEY": "your-api-key-here"
      }
    }
  }
}

Step 4: Test the Integration

Once configured, you can ask your agent questions like:

• "Search the web for the latest AI research papers from 2026"
• "Find Italian restaurants near San Francisco"
• "What are the current top technology news stories?"
• "Search for recent developments in quantum computing"

Available Tools

The Brave Search MCP Server exposes two primary tools:

• brave_web_search — Executes web searches with parameters for query, count (max 20), and offset (max 9).
• brave_local_search — Searches for local businesses with name, address, phone, rating, and hours. Automatically falls back to web search if no local results found.

Conclusion

The Brave Search MCP Server is one of the easiest ways to give your AI agent live web access. With Docker deployment, a generous free API tier, and seamless MCP integration, you can have your agent searching the web in minutes. Visit the Run This Ai tool page for the Docker Compose configuration and complete system requirements.