# Exa MCP

> Lets your agent search the web, read pages as clean text and run multi-step research, through Exa's own hosted MCP server.

Exa's own MCP server connects your agent to Exa's web search. It searches the web and brings back clean, ready-to-use page content rather than a list of links, reads the pages you point it at, and, once you sign in, runs Exa Agent for research that takes more than one search. The Claude plugin and Gemini extension use the same hosted server, and the plugin adds Exa's search and Exa Agent skills.

What you can do with it
- Search the web in natural language and get the content of the best pages back
- Read one or more pages in full as clean Markdown: docs, changelogs, issues, articles and research
- Look up current documentation, release notes and real-world examples while your agent works in your code
- Run Exa Agent for multi-step research, list-building and enrichment, with sources and, if you ask, structured JSON
- Turn on advanced search for domain and category filters, date ranges, highlights, summaries and subpage crawling

How it connects
Exa hosts it at https://mcp.exa.ai/mcp. It works without an API key at free, rate-limited usage. For your Exa team's plan and limits, sign in with OAuth (most agents prompt you; https://mcp.exa.ai/mcp?login asks for it) or send an Exa API key in the x-api-key header. In Claude Code, install the plugin with claude plugin install exa@claude-plugins-official. Exa is also in Claude's connector directory and ChatGPT's plugin directory.

What the plugin's skills add
- search: plans deep research, lead generation, literature reviews and competitive analysis across several searches, then compiles the result
- Exa Agent: guides research runs, enrichment and structured output, including Exa Connect data providers

Good to know
- web_search_exa and web_fetch_exa are on by default, and agent_run joins them once you sign in or add a key. Add ?tools= to the address to choose tools; the list replaces the defaults, so include every tool you want.
- Exa Agent runs are usage-based. A run that outlasts one call reports that it's still running and picks up where it left off.
- Exa also publishes an npm package, exa-mcp-server, that runs it locally with your API key.

- Type: Plugin
- Made by: Exa
- Category: Web & browser
- Formats: Remote MCP, Claude plugins, Extensions
- Version: latest
- Website: https://exa.ai/docs/get-started/exa-mcp
- Source: https://github.com/exa-labs/exa-mcp-server
- Page: https://pluginboard.ai/plugins/exa-mcp

## Works with

- ChatGPT
- Claude
- Muse
- Grok Bot
- Gemini
- Grok
- GitHub Copilot
- Cursor
- Claude Code
- Codex
- Antigravity
- Gemini CLI
- Replit
- OpenClaw
- OpenCode
- Hermes Agent

## Install

### ChatGPT

In ChatGPT on the web: Plugins → + → Add custom MCP server, then paste https://mcp.exa.ai/mcp under Connection. Business and Enterprise plans; Pro can add read-only servers.

### Claude

In Claude (web or desktop): Customize → Plugins → Add → Add marketplace, enter anthropics/knowledge-work-plugins, then add exa

### Muse

In Muse: In a Muse chat, ask it to create a custom connector for the MCP server at https://mcp.exa.ai/mcp. Then tap Connect on the card it shows and sign in, or add your key.

### Grok Bot

In Grok Bot: In a chat with your Bot, say "Add this MCP server: https://mcp.exa.ai/mcp". It confirms the details with you and sets it up, and the tools are ready from your next message.

### Gemini

In Gemini: At gemini.google.com: Settings → Connected Apps → Add a custom app, then paste https://mcp.exa.ai/mcp

### Grok

In Grok on the web: At grok.com/connectors: New Connector → Custom, then enter https://mcp.exa.ai/mcp and sign in if it asks.

### GitHub Copilot

Add to VS Code: vscode:mcp/install?%7B%22name%22%3A%22exa-mcp%22%2C%22type%22%3A%22http%22%2C%22url%22%3A%22https%3A%2F%2Fmcp.exa.ai%2Fmcp%22%7D

Copilot CLI:

```shell
copilot mcp add --transport http exa-mcp https://mcp.exa.ai/mcp
```

### Cursor

Add to Cursor: cursor://anysphere.cursor-deeplink/mcp/install?name=exa-mcp&config=eyJ1cmwiOiJodHRwczovL21jcC5leGEuYWkvbWNwIn0%3D

Add to .cursor/mcp.json:

```json
{
  "mcpServers": {
    "exa-mcp": {
      "url": "https://mcp.exa.ai/mcp"
    }
  }
}
```

### Claude Code

In Claude Code:

```text
/plugin marketplace add anthropics/knowledge-work-plugins
/plugin install exa@knowledge-work-plugins
```

### Codex

Run in your terminal:

```shell
codex mcp add exa-mcp --url https://mcp.exa.ai/mcp
```

### Antigravity

Add to ~/.gemini/config/mcp_config.json:

```json
{
  "mcpServers": {
    "exa-mcp": {
      "url": "https://mcp.exa.ai/mcp"
    }
  }
}
```

### Gemini CLI

Run in your terminal:

```shell
gemini extensions install https://github.com/exa-labs/exa-mcp-server
```

### Replit

In Replit: At replit.com/integrations, open MCP Servers → + Add MCP server, paste https://mcp.exa.ai/mcp (add any headers under Advanced settings), then click Test & save.

### OpenClaw

Run in your terminal:

```shell
openclaw mcp add exa-mcp --url https://mcp.exa.ai/mcp --transport streamable-http
```

If it needs a key or sign-in: For https://mcp.exa.ai/mcp, add --header NAME=value for each header it needs, or --auth oauth and then run openclaw mcp login exa-mcp.

### OpenCode

Run in your terminal:

```shell
opencode mcp add exa-mcp --url https://mcp.exa.ai/mcp
```

If it needs a key: For https://mcp.exa.ai/mcp, add --header 'NAME=value' for each header it needs. Add --global to use it in every project.

### Hermes Agent

Run in your terminal:

```shell
hermes mcp add exa-mcp --url https://mcp.exa.ai/mcp
```

Add to ~/.hermes/config.yaml:

```json
{
  "mcp_servers": {
    "exa-mcp": {
      "url": "https://mcp.exa.ai/mcp"
    }
  }
}
```

If it signs in with OAuth: If https://mcp.exa.ai/mcp asks you to sign in, add --auth oauth to the command, or auth: oauth to its entry.

---
From Pluginboard (https://pluginboard.ai).