Firecrawl MCP Server
Scrape, crawl, map and search the web, returning clean markdown or structured data.
Officialby FirecrawlHosted or on your computerChecked October 18, 2026
Firecrawl's official server turns websites into clean markdown or structured JSON and adds web search, site mapping and full-site crawls. Its hosted endpoint works without signing up at a rate-limited free tier, while an API key or account sign-in adds crawling, the research agent and higher limits.
Who it is for
- Marketers researching competitors
- Developers feeding web data into AI apps
- Analysts collecting data from many pages
Prompts to try
Once it is connected, ask in plain words. Claude decides when to use the server and asks before it changes anything important.
Scrape the pricing pages of these three competitors and put their plans in a comparison table.Crawl our help center and list every page that still mentions the old product name.Search for this week's articles on EU AI rules and summarize the five most useful.How to connect it
In Claude Code
claude mcp add --transport http firecrawl https://mcp.firecrawl.dev/v2/mcp --header "Authorization: Bearer YOUR_API_KEY"
In the Claude desktop app
Customize > Connectors > Discover: search Firecrawl, then select Connect to Claude
Hosted address: https://mcp.firecrawl.dev/v2/mcp
Good to know
Without a key the same URL offers a rate-limited free tier (scrape, search, parse); use https://mcp.firecrawl.dev/v2/mcp-oauth to sign in instead of using a key.
Unlock the setup steps
Free. Leave your email once to open the setup steps for all 150 servers, and get the best new connectors each month.
How we checked it
Checked the GitHub repo (MIT, last commit 2026-10-17), Firecrawl's MCP docs, the firecrawl-mcp npm package (3.30.3) and its Claude connectors listing; the hosted endpoint answers without a key and the search endpoint returns an OAuth challenge.
Source: docs.firecrawl.dev/mcp-server