# Firecrawl MCP — Web Scraping Server for AI Agents

> Official Firecrawl MCP server for AI agents to scrape, crawl, and extract structured data from any website. Supports batch scraping, search, and markdown extraction. 15,000+ stars.

## Install

Merge the JSON below into your `.mcp.json`:

## Quick Use

Add to your `.mcp.json`:

```json
{
  "mcpServers": {
    "firecrawl": {
      "command": "npx",
      "args": ["-y", "firecrawl-mcp"],
      "env": {
        "FIRECRAWL_API_KEY": "fc-YOUR_KEY_HERE"
      }
    }
  }
}
```

Get your free API key at [firecrawl.dev](https://firecrawl.dev). Restart Claude Code and ask: "Scrape https://example.com and summarize the content."

---

## Intro

Firecrawl MCP is the official Model Context Protocol server for Firecrawl, giving AI agents like Claude Code and Cursor the ability to scrape, crawl, and extract structured data from any website. With 15,000+ GitHub stars on the main project, it handles JavaScript-heavy sites and anti-bot protection, outputting clean markdown that LLMs consume directly. Best for AI developers who need agents to read web content in real time. Works with: Claude Code, Cursor, Windsurf, any MCP client. Setup time: under 2 minutes.

---

## Available Tools

### `firecrawl_scrape`
Scrape a single URL and get clean markdown:
```
"Scrape https://docs.anthropic.com and extract the main content"
```

### `firecrawl_crawl`
Crawl an entire website recursively:
```
"Crawl https://example.com with max depth 3 and return all pages"
```

### `firecrawl_map`
Discover all URLs on a website without scraping content:
```
"Map all URLs on https://example.com"
```

### `firecrawl_search`
Search the web and scrape results:
```
"Search for 'best MCP servers 2026' and scrape the top 5 results"
```

### `firecrawl_extract`
Extract structured data using a schema:
```
"Extract all product names, prices, and ratings from https://store.example.com"
```

### `firecrawl_batch_scrape`
Scrape multiple URLs efficiently in one call.

### Key Stats
- 15,000+ GitHub stars (Firecrawl)
- 6 MCP tools available
- Handles JavaScript rendering
- Anti-bot bypass built-in
- Free tier: 500 pages/month

### FAQ

**Q: What is Firecrawl MCP?**
A: Firecrawl MCP is a Model Context Protocol server that lets AI agents scrape, crawl, and extract data from any website through natural language commands.

**Q: Is Firecrawl MCP free?**
A: The MCP server is open-source. Firecrawl offers a free tier with 500 pages/month. Self-hosting is also available.

**Q: Do I need a Firecrawl API key?**
A: Yes, get a free key at firecrawl.dev. Self-host for unlimited usage.

---

## Source & Thanks

> Created by [Mendable](https://github.com/mendableai). Licensed under AGPL-3.0.
>
> [firecrawl](https://github.com/mendableai/firecrawl) — ⭐ 15,000+

Thanks to the Firecrawl team for bringing web scraping superpowers to AI agents.

---

<!-- ZH -->


## Quick Use

Add the following to `.mcp.json`:

```json
{
  "mcpServers": {
    "firecrawl": {
      "command": "npx",
      "args": ["-y", "firecrawl-mcp"],
      "env": {
        "FIRECRAWL_API_KEY": "fc-your-key"
      }
    }
  }
}
```

Get a free API key at [firecrawl.dev](https://firecrawl.dev). Restart Claude Code to use it.

---

## Intro

Firecrawl MCP is Firecrawl's official MCP server, letting AI agents like Claude Code and Cursor scrape, crawl, and extract structured data from any website. The main project has 15,000+ GitHub stars and supports JavaScript rendering and anti-bot bypass. Ideal for developers who need AI to read live web content.

---

## Source & Thanks

> Created by [Mendable](https://github.com/mendableai). Licensed under AGPL-3.0.
>
> [firecrawl](https://github.com/mendableai/firecrawl) — ⭐ 15,000+


---
Source: https://tokrepo.com/en/workflows/firecrawl-mcp-web-scraping-server-ai-agents-3de63ffe
Author: Firecrawl