Firecrawl MCP Server: Scrape and Crawl with Claude
Firecrawl's MCP server lets Claude search, scrape, crawl, map and interact with websites, returning clean Markdown or JSON. Hosted and npx setup, pros, cons.
By Claudekit Team Published 1 min read
- Type
- MCP Server
- Pricing
- Freemium
- Repository
- github.com
- GitHub stars
- 7,673
- Website
- firecrawl.dev
The Firecrawl MCP Server brings Firecrawl, a web data API for AI agents, into Claude. It can search the web, scrape any URL into clean Markdown or schema-shaped JSON, parse PDFs and documents, crawl or map whole sites, interact with dynamic pages in a real browser (click, type, fill forms), run multi-source research agents, and monitor pages for changes.
Who it is for
Developers and analysts who need more than a single page fetch: competitive research, docs ingestion, structured extraction or change monitoring.
Install
The fastest start is the hosted server, with a keyless free tier for scrape, search and parse:
https://mcp.firecrawl.dev/v2/mcpUse https://mcp.firecrawl.dev/v2/mcp-oauth to connect your account interactively. In Claude Code:
claude mcp add --transport http firecrawl https://mcp.firecrawl.dev/v2/mcp-oauthTo run it locally with an API key, the README’s command is:
env FIRECRAWL_API_KEY=fc-YOUR_API_KEY npx -y firecrawl-mcpA firecrawl plugin is also listed in Anthropic’s official marketplace.
Use cases
- Docs ingestion: crawl a docs site with path filters and summarize the API surface.
- Structured extraction: scrape product pages into JSON matching a schema you provide.
- Gated flows: use
firecrawl_interactto click through a page before reading it. - Monitoring: track a pricing page and get diffs when it changes.
Pros
- Broad toolset covering search, scrape, crawl, browser actions and agents.
- Keyless tier for quick trials; Docker image and self-hosting supported.
- Clear guidance in the README on which tool to use when.
Cons and caveats
- Crawls consume credits fast; always set
limitand path filters. - 27 tools add noticeable context; disable what you do not use.
- Respect site terms and robots rules when scraping.
Links
- Repo: firecrawl/firecrawl-mcp-server
- Related: Exa MCP, Playwright MCP, Brave Search MCP
Frequently asked questions
Can I use Firecrawl MCP without an API key?
Yes, on the hosted keyless free tier: scrape, search and parse work rate-limited. Crawl, map and agent tools need a key or OAuth.
Does it support self-hosted Firecrawl?
Yes. The README lists cloud and self-hosted support; point the server at your own Firecrawl instance through its configuration.
How many tools does it expose?
Up to 27 with the full profile and default settings. Environment flags can turn off feedback tools to shrink the list.