M
MCP Directory CURATED DIRECTORY
911 MCP servers indexed
中文
← Back to directory
E

Exa MCP Server

Official
Exa's official MCP server for web search and crawling built for AI applications
How to install ↓ Compare GitHub source repository ↗
Category
Other #5 of 230
Stars
★ 5.1k Very popular
Transport
stdio (local process) · SSE
Runtime
Hosted (no local runtime) · Node.js
Credentials
API key / credential required
License
MIT
Last commit
Tools
4
73FMRS · B

Officially maintained by Exa; the README clearly distinguishes default-enabled tools from ones requiring manual opt-in, with clear permission boundaries. Both local and remote paths are documented. This review is a static check of source and docs, not an actual call test.

Strongest · Setup experience 16/20 Weakest · Reliability 12/20

Reliability
12/20
Security and permissions
13/20
Maintenance
16/20
Documentation
16/20
Setup experience
16/20

Test out of dateListed 2026-07-24

Read the FMRS scoring method →

Fit and risk

What it can accessUses the network

Best for

  • Q&A and research scenarios that need real-time, clean web content
  • Users with an existing Exa account who want a zero-deployment remote option

Not for

  • Scenarios requiring a strict allowlist of search sources that can't use a third-party search service
  • Budget-sensitive, high-volume search scenarios (remote/API calls may incur cost)

Required permissions

  • Requires an Exa API key for full functionality; OAuth login is optional in remote mode
  • web_fetch_exa retrieves content from a specified URL; agent_run performs multi-step autonomous actions and needs the user to have clear expectations of Agent behavior

Risks and side effects

  • The API key should be stored carefully — avoid leaving it in plaintext in client config files on untrusted machines
  • agent_run is a multi-step autonomous tool with a longer action chain — try it on low-risk tasks first

Setup

Before you start

Runtime：Hosted (no local runtime) · Node.js

EXA_API_KEY requiredsecret Exa API key, obtained at https://dashboard.exa.ai/api-keys; only needed when running via the npm package locally.

1) Local: npx -y exa-mcp-server, after getting an EXA_API_KEY from https://dashboard.exa.ai/api-keys and setting it as an env var.
2) Remote: point your client at https://mcp.exa.ai/mcp, select enabled tools via the URL's tools parameter (e.g. ?tools=web_search_exa,web_fetch_exa), or add ?login to trigger an OAuth login.

claude_desktop_config.json 
{
  "mcpServers": {
    "exa": {
      "command": "npx",
      "args": ["-y", "exa-mcp-server"],
      "env": {
        "EXA_API_KEY": "<YOUR_EXA_API_KEY>"
      }
    }
  }
}

Shown for Claude Desktop. Other clients may use a different file or key (VS Code uses "servers") — the configurator below converts it.

.vscode/mcp.json
{
  "servers": {
    "exa": {
      "command": "npx",
      "args": [
        "-y",
        "exa-mcp-server"
      ],
      "env": {
        "EXA_API_KEY": "<YOUR_EXA_API_KEY>"
      }
    }
  }
}

Goes in your project's .vscode/mcp.json (VS Code uses a "servers" key).

Terminal
claude mcp add exa -e 'EXA_API_KEY=<YOUR_EXA_API_KEY>' -- npx -y exa-mcp-server

Run it in a terminal; replace any <…> placeholders with your own values first.

Generate and validate client config

Check that it works

Check that web_search_exa and web_fetch_exa appear in your client's tool list; ask the AI to search a current topic and return source URLs to confirm the connection works.

Troubleshooting

  1. No search results: check that EXA_API_KEY is set correctly and the account has remaining quota
  2. Remote endpoint connection fails: confirm network access to mcp.exa.ai and that the URL's tools parameter is spelled correctly
  3. Local npm install fails: confirm the Node.js version meets requirements

Things to try

Once connected, you can ask your AI assistant things like:

  • Search the web for recent papers on transformer attention efficiency and give me arXiv links
  • Fetch the full content of https://docs.exa.ai/reference/exa-mcp
  • Use advanced search to find news about AI safety funding since 2024
  • Run an Exa Agent to build a list of AI infrastructure startups in San Francisco

Tools 4

web_search_exa read-only
Search the web for any topic and get clean, ready-to-use content (enabled by default)
web_fetch_exa read-only
Get the full content of a specific webpage from a known URL (enabled by default)
web_search_advanced_exa read-only
Advanced search with filters for domains, dates, and content options (off by default)
agent_run read-only
Run an Exa Agent for multi-step research, list-building, or structured extraction (requires OAuth or API key)

Use cases

Have an AI assistant search the live web for current information
Fetch the full content of a specific URL for summarization or analysis
Filter search results by domain or date range for research
Enable Exa Agent for multi-step information-gathering tasks

Supported clients

Claude Desktop
Claude Code
VS Code
Cursor
Windsurf
Zed
Amp
Kiro
LM Studio
Antigravity

Listed from the project's documentation, not tested by this site.

Overview

Officially maintained by Exa, this MCP server exposes Exa's search engine (purpose-built for AI applications) — web search, page fetching, and advanced filtering — to AI assistants. Runs locally via an npm package, or connects directly to Exa's hosted SSE remote endpoint with tools selected via URL parameters.

Similar servers

BioMCP 75 · B

One binary. One grammar. Evidence from the biomedical sources you already trust.

★ 659 · Tools 11 Compare with this →
Compare with other MCP servers

Source revision e9c3b0126a33 Data synced 2026-10-19 Read the FMRS scoring method