mcp-registry/exa-mcp-server

    ==================
      
       /// MCP ///
      /// EXA ///
        
    ==================
        
    [server:online]
    [protocol:ready]

exa-mcp-server

by exa-labs

Model Context Protocol (MCP) server that lets AI assistants (e.g., Claude Desktop) access Exa AI’s web-search API with optional tools for company research, crawling, LinkedIn search and deep research automation.

2.0k
174
Enterprise

01

web_search_exa

Performs real-time web searches with optimized results and content extraction.

02

company_research

Comprehensive company research tool that crawls company websites to gather detailed information about businesses.

03

crawling

Extracts content from specific URLs, useful for reading articles, PDFs, or any web page when you have the exact URL.

04

linkedin_search

Search LinkedIn for companies and people using Exa AI. Include company names, person names, or specific LinkedIn URLs in your query.

05

deep_researcher_start

Start a smart AI researcher for complex questions, which searches the web, reviews many sources, and generates a detailed research report.

06

deep_researcher_check

Check whether a research task is complete and retrieve the comprehensive report.

Installation

1. Prerequisites
• Node.js ≥ 18 and npm (or pnpm/yarn)
• An Exa API key with MCP access
2. Clone the repository
git clone https://github.com/exa-labs/exa-mcp-server.git cd exa-mcp-server
3. Install dependencies
npm install # or: pnpm install / yarn install
4. Configure environment
• Copy the example env file and fill in required values:
cp .env.example .env # open .env and set # EXA_API_KEY=<your-key> # PORT=3000 # optional override of default port # LOG_LEVEL=info # trace | debug | info | warn | error
5. Build TypeScript → JavaScript (optional in dev mode)
npm run build # transpiles src/**/*.ts to dist/
6. Run the MCP server
# Development (auto-reload) npm run dev # Production npm start # equivalent to: node dist/index.js
7. Docker (alternative)
docker build -t exa-mcp . docker run -e EXA_API_KEY=<your-key> -p 3000:3000 exa-mcp
The service will be reachable at http://localhost:3000 by default.

Documentation

License: MIT License
Updated 7/30/2025

Table of Contents