Brave Search MCP Server

What is Brave Search MCP Server?

An MCP server implementation that integrates the Brave Search API, providing both web and local search capabilities. It is designed for developers and LLM users who need real-time search results with flexible filtering and smart fallbacks.

How to use Brave Search MCP Server?

Obtain a Brave Search API key from the Brave developer dashboard, then configure the server via an env variable BRAVE_API_KEY. It can be run using Docker (mcp/brave-search) or NPX (@modelcontextprotocol/server-brave-search). Configuration examples are provided for Claude Desktop (claude_desktop_config.json) and VS Code (User Settings or .vscode/mcp.json), including one-click install buttons.

Key features of Brave Search MCP Server

• Web and local search in a single server
• Pagination and freshness controls for web queries
• Local search with business details and automatic fallback
• Flexible filtering: result types, safety levels, content freshness
• Max 20 results per request (configurable count)

Use cases of Brave Search MCP Server

• General web searches with pagination and date filtering
• Finding local businesses, restaurants, and services with detailed information
• News and article searches with freshness controls
• Integrating Brave Search into AI assistants like Claude Desktop or VS Code

FAQ from Brave Search MCP Server

How do I get a Brave Search API key?

Sign up for a Brave Search API account at their site, choose a plan (Free tier available with 2,000 queries/month), and generate an API key from the developer dashboard.

What tools does the server provide?

Two tools: brave_web_search (web searches with pagination, filtering, freshness) and brave_local_search (local business/service searches with automatic fallback to web if no local results).

What are the maximum result counts per request?

Both tools accept an optional count parameter, with a maximum of 20 results per call.

Does the server handle cases where local search returns nothing?

Yes. Local search automatically falls back to web search when no local results are found.

What runtime or environment does the server require?

It can run via Docker (image mcp/brave-search) or directly via NPX with Node.js. No additional runtime dependencies are mentioned beyond the Brave Search API key.