MCP Directory

How to add Firecrawl to Claude Desktop

Official Firecrawl MCP server — scrape, crawl, map, search, and structured extraction for any LLM client. Paste the config into ~/Library/Application Support/Claude/claude_desktop_config.json and restart Claude Desktop.

Last updated June 15, 2026 · 6.6k★ · stdio · apikey · official

Claude Desktop config for Firecrawl

npx -y firecrawl-mcp
{
  "mcpServers": {
    "firecrawl": {
      "command": "npx",
      "args": [
        "-y",
        "firecrawl-mcp"
      ],
      "env": {
        "FIRECRAWL_API_KEY": "<your-firecrawl-api-key>"
      }
    }
  }
}

Setup steps

  1. 1Open Claude Desktop → Settings → Developer → Edit Config (this opens ~/Library/Application Support/Claude/claude_desktop_config.json).
  2. 2Paste the Firecrawl config below under the top-level "mcpServers" key.
  3. 3Fill in any placeholder secrets (API keys, paths) in the snippet.
  4. 4Save the file, then fully quit and reopen Claude Desktop.
  5. 5Open a chat and confirm Firecrawl's tools appear under the 🔌 tools menu.

Before you start

  • Node.js (18+) with npx, or install globally via npm
  • A Firecrawl API key (format fc-...) from https://www.firecrawl.dev/app/api-keys
  • Optional: FIRECRAWL_API_URL to target a self-hosted Firecrawl instance instead of the cloud API

What Firecrawl can do in Claude Desktop

firecrawl_scrape

Scrape a single URL and return clean Markdown or JSON content.

firecrawl_batch_scrape

Scrape multiple known URLs in one async batch job.

firecrawl_check_batch_status

Poll the progress and results of a batch scrape job.

firecrawl_map

Discover and list all indexed/linked URLs on a website.

firecrawl_search

Run a web search and optionally scrape the result pages' content.

firecrawl_crawl

Asynchronously crawl a site across multiple pages with depth and limit controls.

firecrawl_check_crawl_status

Poll the status and accumulated results of a crawl job.

firecrawl_extract

Extract structured data from pages using an LLM and a custom JSON schema/prompt.

Security

Requires a FIRECRAWL_API_KEY that gates billable credits; keep it secret and watch for credit-usage warnings. The crawl/scrape tools fetch arbitrary URLs you supply, so treat all returned page content as untrusted model input.

Firecrawl + Claude Desktop FAQ

Where is the Claude Desktop config file?

Claude Desktop reads MCP servers from ~/Library/Application Support/Claude/claude_desktop_config.json. Paste the Firecrawl config there under the "mcpServers" key and restart the client.

Is Firecrawl safe to use with Claude Desktop?

Requires a FIRECRAWL_API_KEY that gates billable credits; keep it secret and watch for credit-usage warnings. The crawl/scrape tools fetch arbitrary URLs you supply, so treat all returned page content as untrusted model input.

Do I need an API key?

Yes. The server is a client for the hosted Firecrawl API and requires a FIRECRAWL_API_KEY (format fc-...). You can self-host the Firecrawl backend and point FIRECRAWL_API_URL at it.

Is there a free tier?

Firecrawl offers a free plan with a limited monthly credit allowance; scrapes, crawls, and extractions consume credits. Heavy crawling requires a paid plan.

Which MCP clients does it support?

Any MCP-compatible client over stdio, including Claude Desktop, Cursor, VS Code, and Windsurf. It can also run as an HTTP Streamable or local SSE server for remote setups.

View repo Full Firecrawl page