MCP servers / Firecrawl

Firecrawl MCP server: setup for Claude Code, Claude Desktop and Cursor

Firecrawl is a web scraping service, and this server connects your AI to it. Beyond scraping a single page, it can map a whole site, crawl many pages, search the web, and run longer research jobs.

It is much bigger than "scrape a page": the server we tested offers 27 tools, including live browser sessions, scheduled monitors, and search tools for research papers and public GitHub issues.

Runs with

npx (Node.js)

Package

firecrawl-mcp on npm

Needs

Node.js (for npx) and a Firecrawl API key from firecrawl.dev.

Tools

27 (16 read-only, 11 can change things)

Source

firecrawl/firecrawl-mcp-server

Last checked

2026-09-20 (server reported version 3.24.0)

Config for Firecrawl

Generated by the same code as the free builder, so it matches what you get there. Pick your client, then copy.

Claude Code

Save as .mcp.json in your project root. Claude Code asks you to approve it the first time.

{
  "mcpServers": {
    "firecrawl": {
      "command": "npx",
      "args": [
        "-y",
        "firecrawl-mcp"
      ],
      "env": {
        "FIRECRAWL_API_KEY": "${FIRECRAWL_API_KEY}"
      }
    }
  }
}

Set FIRECRAWL_API_KEY in your environment before starting Claude Code. The config only references the variable, so it is safe to commit.

Or add it from a terminal instead (single quotes work in bash, zsh and PowerShell; more on this command):

claude mcp add firecrawl -e 'FIRECRAWL_API_KEY=${FIRECRAWL_API_KEY}' -- npx -y firecrawl-mcp

Claude Desktop

Merge into claude_desktop_config.json (Settings → Developer → Edit Config), then fully quit and restart Claude Desktop.

{
  "mcpServers": {
    "firecrawl": {
      "command": "npx",
      "args": [
        "-y",
        "firecrawl-mcp"
      ],
      "env": {
        "FIRECRAWL_API_KEY": "paste-your-key-here"
      }
    }
  }
}

Claude Desktop keeps FIRECRAWL_API_KEY in this file as plain text. Keep the file private and never share or commit it.

Cursor

Save as .cursor/mcp.json in your project, or ~/.cursor/mcp.json for all projects.

{
  "mcpServers": {
    "firecrawl": {
      "command": "npx",
      "args": [
        "-y",
        "firecrawl-mcp"
      ],
      "env": {
        "FIRECRAWL_API_KEY": "${env:FIRECRAWL_API_KEY}"
      }
    }
  }
}

Set FIRECRAWL_API_KEY in your environment before launching Cursor. The config only references the variable, so it is safe to commit.

What it can do

These are all 27 tools the server reported when we started it and asked on 2026-09-20. The labels come from the hints each tool declares about itself.

• firecrawl_scrape can change things
  Retrieve and extract content from one supplied URL through Firecrawl.
• firecrawl_map read-only
  Enumerate URLs indexed under one website through Firecrawl without fetching each page's content.
• firecrawl_search read-only
  Search web, news, or image sources and return ranked results.
• firecrawl_search_feedback can change things
  Records schema-validated quality feedback for a prior `firecrawl_search` UUID `searchId`.
• firecrawl_feedback can change things
  Submit concise quality feedback for a completed search, scrape, parse, or map job.
• firecrawl_crawl can change things
  Start a multi-page crawl at a website URL, poll it to a terminal state, and return the final status and collected data.
• firecrawl_check_crawl_status read-only
  Retrieve the current status, progress, and available results for an existing crawl ID.
• firecrawl_extract read-only
  Deprecated compatibility entry point.
• firecrawl_agent can change things
  Start an asynchronous web research job from a prompt, optional seed URLs, and an optional JSON schema.
• firecrawl_agent_status read-only
  Retrieve progress or final results for a `firecrawl_agent` job ID.
• firecrawl_interact can change things
  Open or reuse a live browser session to navigate a page, click controls, fill fields, or run browser code.
• firecrawl_interact_stop can change things
  Stop the live interact session associated with a `scrapeId` and release its resources.
• firecrawl_parse read-only
  Parse one supported document into markdown, HTML, links, summary, targeted answers, or JSON matching a schema.
• firecrawl_monitor_create can change things
  Create a recurring scrape, crawl, or search monitor that compares each check with its retained predecessor.
• firecrawl_monitor_list read-only
  List monitors for the authenticated account with optional pagination controls.
• firecrawl_monitor_get read-only
  Retrieve one monitor by ID, including its configuration and current state.
• firecrawl_monitor_update can change things
  Patch an existing monitor by ID.
• firecrawl_monitor_delete can change things
  Permanently delete a monitor by ID and stop its future schedule.
• firecrawl_monitor_run can change things
  Queue an immediate check for a monitor outside its normal schedule.
• firecrawl_monitor_checks read-only
  List historical checks for a monitor, optionally filtered by status and bounded by a result limit.
• firecrawl_monitor_check read-only
  Retrieve one monitor check and its page-level results, optionally filtered by page status.
• firecrawl_research_search_papers read-only
  Search paper metadata and abstracts with a natural-language query across the indexed corpus, which spans biomedical, life-science, and clinical literature (PubMed,…
• firecrawl_research_inspect_paper read-only
  Retrieve canonical metadata for one paper ID, such as an arXiv, PMC, PMID, or DOI identifier.
• firecrawl_research_related_papers read-only
  Find citation-graph candidates from one to ten `seed_ids`; the first ID is the primary seed and later IDs are anchors.
• firecrawl_research_read_paper read-only
  Retrieve in-body passages from one paper that are relevant to a specific question.
• firecrawl_research_search_github read-only
  Search indexed public GitHub issue, pull-request, and README content.
• firecrawl_developer_search read-only
  For a developer question — code behaviour, a library or framework, an API contract, an error message, or a known bug — search an index built for coding agents.

Watch out for

• Requests run on Firecrawl's service under your account, so check your plan's limits before letting the AI crawl freely.
• The monitor tools create recurring checks that run on a schedule, so they can keep running after your chat ends. Check your Firecrawl account for anything the AI set up.
• The server lists firecrawl_extract as deprecated in its own description, so expect the tool list to change between versions.
• You can point it at a self-hosted Firecrawl with FIRECRAWL_API_URL (from the README), in which case the API key is optional.

Try asking

• “Map the docs site at this URL and list its main sections.”
• “Scrape this page and extract the product names and prices as a table.”
• “Search the web for recent posts about this error message and summarize the common fixes.”

Related servers

• Fetch: Download a web page and convert it to readable text.
• Playwright: Drive a real browser to test pages and take screenshots.