Exa

You connect your AI assistants to Exa’s MCP server to unlock real-time web search, research tooling, and fast code context. This enables your agents to fetch live information, explore open sources, and get code snippets without hallucination, directly within your conversational flows.

How to use

Use the Exa MCP server to power your MCP client. You will configure a connection, enable the tools you need, and then run your MCP-enabled assistant with the server in place. When you query your assistant, it can perform web searches, read source docs, and fetch code examples through the available tools.

Key tools include the code search capability for finding up-to-date code snippets, and real-time web research tools. You can customize which tools are active to match your workflow. If you want to limit tool usage, specify only the tools you need when configuring your MCP client.

How to install

Prerequisites: Node.js (install from nodejs.org) and npm (comes with Node.js). You may also install via npm or npx depending on your preferred workflow.

Install the MCP server globally so you can run it from anywhere.

npm install -g exa-mcp-server

Additional sections

Configuration is done by creating MCP server entries that describe how to connect to the Exa MCP server. You can connect directly to the hosted server via HTTP or run a local stdio-based instance that your MCP client can start.

{
  "mcpServers": {
    "exa": {
      "type": "http",
      "url": "https://mcp.exa.ai/mcp",
      "headers": {
        "Remove-Me": "Disable web_search_exa tool if you're just coding. To 100% call exa-code, say 'use exa-code'."
      }
    }
  }
}

Security and keys

Secure your API key by passing it via environment variables or secure URL parameters as shown in the usage examples. Do not expose keys in client-side code or publicly readable configuration files.

Available tools

get_code_context_exa

Search and retrieve relevant code snippets, examples, and documentation from open source repositories and docs to help you code faster and with accurate context.

web_search_exa

Perform real-time web searches with optimized results and content extraction for up-to-date information.

company_research

Crawl company websites to gather detailed information about businesses.

crawling

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

linkedin_search

Search LinkedIn for companies and people using Exa AI, given names or URLs.

deep_researcher_start

Start a smart AI researcher task that surveys the web and sources to create a detailed research report.

deep_researcher_check

Check progress and obtain the final report for a started deep research task.