Last Verified: September 2026 — commands checked against the official Firecrawl docs and firecrawl-mcp v3.25.3; keyless endpoint tested live
You ask Claude Code to pull the plans from a pricing page, and it hands back a vague summary instead of the table. Its built-in WebFetch doesn’t run JavaScript, and the agent gets a processed summary rather than the page itself. JS-rendered docs, pricing tables, and “read these 30 pages and compare them” jobs are where it falls apart.
Firecrawl MCP fixes that with one command. It gives the agent a proper web toolkit (~7.5k GitHub stars, maintained by the Firecrawl team). It can search the web, scrape any URL into clean markdown or JSON, map every URL on a site, crawl whole doc sections, and parse PDFs. The part most guides miss is that there’s now a hosted keyless endpoint. Search, scrape, and parse work with no API key, no account, and no Node.js install.
Only need library docs? Our Context7 MCP setup is the better tool for that. Firecrawl is for everything else on the web.
Quick Start
One command. No API key, no Node.js:
claude mcp add --transport http --scope user firecrawl https://mcp.firecrawl.dev/v2/mcpRestart Claude Code, then ask: “Use Firecrawl to scrape https://docs.firecrawl.dev and summarize it”. Total time: about 1 minute.
⚡ Copy This Prompt: Let Claude Code Install and Verify It For You
Skip the manual steps below and hand the whole job to the agent instead:
You have shell access and the `claude mcp` CLI on this machine. Set up the Firecrawl MCP server (hosted, keyless) and report back. Do not tell me it's done unless step 4 actually confirms it:
1. Run `claude mcp list`. If a server named firecrawl already exists, stop and tell me how it's configured instead of adding a second one.
2. Add the hosted keyless server (no API key, no Node.js needed):
`claude mcp add --transport http --scope user firecrawl https://mcp.firecrawl.dev/v2/mcp`
3. Run `claude mcp list` again and confirm firecrawl shows as Connected.
4. Tell me to restart Claude Code, then after restart use firecrawl_scrape on https://docs.firecrawl.dev and show me the first 10 lines of the markdown it returns.
Never ask me to paste an API key into this chat. If I want the full tool set later, I'll switch to the sign-in URL myself.
If any step fails, give me the exact error and the command that produced it. Do not skip ahead.The prompt deliberately keeps your API key out of the chat. Firecrawl’s own docs say never to put a key in an agent conversation. The sign-in URL in Step 3 below gets you the full tool set without ever handling a key.
What You’ll Need
| Requirement | Why You Need It | Time |
|---|---|---|
| Claude Code installed | Provides the claude mcp command used to register the server | 0 min |
| Nothing else (keyless) | The hosted endpoint runs on Firecrawl’s servers. Search, scrape, and parse are free, rate-limited per IP per day | 0 min |
| Free Firecrawl account (optional) | Unlocks crawl, map, agent, and the rest of the tools. The free plan includes 1,000 credits, no card | ~2 min |
| Node.js 22+ (local mode only) | Only if you run the server on your own machine or point it at a self-hosted Firecrawl | 0–3 min |
Step-by-Step Setup
Prefer to do it by hand instead of delegating to the agent? Here’s the manual version.
Step 1 — Add the keyless server
~1 min
--scope user makes it available in every project:
claude mcp add --transport http --scope user firecrawl https://mcp.firecrawl.dev/v2/mcpThis is Firecrawl’s documented Claude Code command with --scope user added. Keyless mode exposes three tools: firecrawl_search, firecrawl_scrape, and firecrawl_parse.
Step 2 — Restart and run a first test
~1 min
Restart Claude Code so it loads the new tools, then try:
Search the web for the latest Firecrawl release notes and summarize the sources.That’s Firecrawl’s own suggested first prompt. You should see firecrawl_search in the tool calls.
Step 3 — Unlock the full tool set with sign-in (optional)
~2 min
Crawl, map, agent, and the other tools need an account. The easiest route is browser sign-in, with no API key to copy. Remove the keyless entry first, since Firecrawl’s docs say not to add a second Firecrawl server:
claude mcp remove firecrawlclaude mcp add --transport http --scope user firecrawl https://mcp.firecrawl.dev/v2/mcp-oauthThen run /mcp inside Claude Code and complete the sign-in in your browser. You can review or revoke the connection later in Firecrawl’s MCP settings.
Step 4 — Or run it locally with an API key (optional)
~3 min
Use this if you’d rather run the open-source server yourself, or point it at a self-hosted Firecrawl via FIRECRAWL_API_URL. It needs Node.js 22+ and a key from your Firecrawl dashboard. Type the key into your terminal, not into a chat:
claude mcp add firecrawl --scope user -e FIRECRAWL_API_KEY=fc-YOUR_API_KEY -- npx -y firecrawl-mcpWindows:
claude mcp add firecrawl --scope user -e FIRECRAWL_API_KEY=fc-YOUR_API_KEY -- cmd /c npx -y firecrawl-mcpRunning n8n? Start the server with HTTP_STREAMABLE_SERVER=true and point n8n’s MCP client at http://localhost:3000/mcp.
What Each Piece Does
| Tool | What the Agent Uses It For | Keyless? |
|---|---|---|
firecrawl_search | Web search with ranked results, optionally fetching page content in the same call | Yes |
firecrawl_scrape | One known URL → clean markdown, or JSON matching a schema you give it | Yes |
firecrawl_parse | PDFs, Word docs, spreadsheets, HTML files → markdown or JSON | Yes |
firecrawl_map | List every URL on a site without fetching content, which is handy before a crawl | No |
firecrawl_crawl | Pull content from many pages under a site, bounded by limit and include/exclude paths | No |
firecrawl_interact | Click, type, or navigate on a page before reading it | No |
firecrawl_agent | Autonomous multi-source research that returns structured data | No |
firecrawl_monitor_* | Check the same page on a schedule and get diffs when it changes | No |
firecrawl_credit_usage | Check remaining credits and monthly consumption | No |
With sign-in or an API key, the full profile registers 26 tools, including developer search over GitHub issues and PRs, research-paper search, and crawl/agent status checks. The table covers the ones you’ll actually reach for.
Verify It’s Actually Working
claude mcp list! firecrawl Needs authentication → sign-in URL: run /mcp and finish the browser sign-in
✘ firecrawl 429 Too Many Requests → keyless daily limit hit; sign in or add a key
“Connected” only confirms the server answered. The real test is a page WebFetch struggles with, such as a JavaScript-heavy page:
Use firecrawl_scrape on [A JS-HEAVY PAGE, e.g. a pricing page] and give me the exact plan names and prices as a markdown table. Quote them verbatim from the scraped content, don't summarize.If the tool calls show firecrawl_scrape and the table matches what you see in your browser, it’s working end to end.
Common Mistakes to Avoid
- Opening the MCP URL in a browser.
https://mcp.firecrawl.dev/v2/mcpis a server endpoint for your MCP client, not a web page. Sign-in happens through/mcpin Claude Code. - Pasting your API key into the chat. Firecrawl’s docs are explicit: never put a key in an agent conversation or in the server URL. Use browser sign-in, or pass it with
-e/-Hin your own terminal. - Adding a second Firecrawl entry to “upgrade”. Remove the keyless server before adding the sign-in one, or Claude Code ends up with duplicate tools.
- Expecting crawl or map on keyless. Keyless is limited to search, scrape, and parse. If
firecrawl_crawlis missing, that’s why, not a broken install. - Crawling without a limit. Every page costs credits. Run
firecrawl_mapfirst, then crawl withlimitandincludePathsset so a docs crawl doesn’t eat your whole free allowance.
Q&A
Is it really free without an API key?
Yes. The hosted keyless endpoint offers search, scrape, and parse for free, capped per IP address per day by both a request limit and a credit limit. Past that you get a 429. A free account adds 1,000 credits and the full tool set.
How is this different from Claude Code’s built-in WebFetch?
WebFetch fetches a single URL and hands the agent a processed summary. It doesn’t run JavaScript. Firecrawl renders the page and returns the actual content as markdown or structured JSON. It can also search, map, and crawl many pages, which WebFetch can’t.
Do I need Node.js?
Not for the hosted endpoint. Node.js 22+ is only needed if you run firecrawl-mcp locally, for example against a self-hosted Firecrawl.
Can I use it in n8n?
Yes. Run the server locally with HTTP_STREAMABLE_SERVER=true and connect n8n to http://localhost:3000/mcp. curl http://localhost:3000/health should return ok.
How do I remove it?
claude mcp remove firecrawl. If you used sign-in, also revoke the connection in Firecrawl’s MCP settings.
Official Resources
- Firecrawl MCP Server on GitHub — source, full tool reference, changelog
- Firecrawl MCP docs — keyless, sign-in, and API-key setup
- Run Firecrawl MCP locally — stdio, Streamable HTTP, self-hosted API
- Rate limits — keyless allowance and plan limits
- Claude Code MCP Documentation — the
claude mcpcommand reference
If this guide got Claude Code reading the real web instead of a summary of it, follow us on X for more setup guides and prompt tips.
Leave a Reply