---
title: "Firecrawl MCP: review, setup & limitations · Undominated.ai"
canonical: https://undominated.ai/mcp-servers/firecrawl/
description: "Provides web scraping, search, crawling and browser-interaction jobs through Firecrawl’s API."
---

# Firecrawl MCP: review, setup & limitations · Undominated.ai

> Provides web scraping, search, crawling and browser-interaction jobs through Firecrawl’s API.

[← Explore all mcp servers](/mcp-servers/)

WEB EXTRACTION AND AUTOMATION / Firecrawl

# Firecrawl MCP

Provides web scraping, search, crawling and browser-interaction jobs through Firecrawl’s API.

 [See setup guidance ↓](#setup)[Original source ↗](https://github.com/firecrawl/firecrawl-mcp-server)

SOURCE REVIEW

 Reviewed 2026-10-13
 Evidence 6 linked sources
 Publisher Firecrawl
 Licence [MIT ↗](https://github.com/firecrawl/firecrawl-mcp-server/blob/1554b675ebb697e2d630d3164bbe7601bb8cdcb0/LICENSE)
 Revision 1554b675ebb6
 [Read what was—and wasn’t—checked ↓](#review)

“Search the web and get full page content”

 [Firecrawl · upstream description ↗](https://github.com/firecrawl/firecrawl-mcp-server/blob/1554b675ebb697e2d630d3164bbe7601bb8cdcb0/README.md) Our analysis follows below.

01 / THE REASONING

## Why this made the selection.

 - Distinguishes simple extraction/search from broader crawl and interaction capabilities.
- Documents hosted keyless, OAuth and API-key modes as well as a local adapter.

### A good fit for

 - Turning permitted web pages into structured context for an assistant.
- Re-checking the same page on a schedule and reading the diffs, with the firecrawl_monitor_* tools.

### Weigh up before choosing

 - Keyless hosted access exposes a smaller tool set; advanced jobs need authentication.
- Interaction tools can operate remote web pages and are broader than read-only fetching.
- The repository moved from github.com/mendableai/firecrawl-mcp-server to github.com/firecrawl/firecrawl-mcp-server (https://api.github.com/repos/mendableai/firecrawl-mcp-server answers HTTP 301 to the repository whose full_name is firecrawl/firecrawl-mcp-server, checked 2026-10-13).
- firecrawl_interact runs a prompt or code turn against a remote browser page and can save cookies and localStorage to a named profile when its session ends.

02 / THE REVIEW RECORD

## What we actually inspected.

Source review has boundaries. A clear record is more useful than a “safe” badge.

### Material inspected

 - README.md (installation, capability and permission sections)
- package.json
- src/index.ts
- LICENSE
- src/service/index.ts

### Our findings

 - The current guidance rejects API keys embedded in URLs and deprecates older credential-in-path routes.
- The implementation includes multiple authentication modes and tool profiles; the full profile should not be presented as a search-only integration.
- The repository was renamed from mendableai/firecrawl-mcp-server to firecrawl/firecrawl-mcp-server; this record now pins the new name. Evidence: https://api.github.com/repos/mendableai/firecrawl-mcp-server.
- At this commit src/index.ts delegates to server.ts and service/index.ts; service/index.ts marks the credential-in-path route deprecated, and the README says never to put an API key in the server URL.
- The README’s tool sections cover firecrawl_scrape, firecrawl_map, firecrawl_search, firecrawl_search_feedback, firecrawl_feedback, firecrawl_crawl, firecrawl_check_crawl_status, firecrawl_parse, firecrawl_agent, firecrawl_agent_status, firecrawl_interact, firecrawl_interact_stop, the firecrawl_research_* and firecrawl_monitor_* families, firecrawl_developer_search, firecrawl_gov_search, firecrawl_find_tools and firecrawl_credit_usage. It says the full profile lists 27 tools with default settings.
- The README now describes the hosted /v2/mcp-search endpoint as a fixed-scope surface of nine tools, including firecrawl_find_tools and firecrawl_scrape. At the previous pin it described six read-only tools.
- The README’s Claude Desktop block launches npx -y firecrawl-mcp with FIRECRAWL_API_KEY set to a placeholder. package.json still names the package firecrawl-mcp.

### Not established by this review

 - Server startup, authentication and tool execution were not tested.

The review applies to the material and revision named here. A newer upstream release can change its behavior.

03 / PUT IT TO WORK

## Connect a server deliberately.

[Upstream setup instructions ↗](https://github.com/firecrawl/firecrawl-mcp-server/blob/1554b675ebb697e2d630d3164bbe7601bb8cdcb0/README.md)

DOCUMENTED COMMAND

 npx -y firecrawl-mcp Copy command ↗

Copying does not execute this command. It may retrieve a newer version than the reviewed source.

 - Configure the documented Firecrawl API credentials or choose a supported keyless/hosted setup.
- Review enabled crawl, agent and interaction tools against the intended websites and task.

### Before you start

 - Node.js for the local adapter; a Firecrawl account for authenticated jobs.

### Documented configuration example

Use the client and authentication context described in the setup guidance. Configuration formats vary between hosts.

 {
 "mcpServers": {
 "mcp-server-firecrawl": {
 "command": "npx",
 "args": [
 "-y",
 "firecrawl-mcp"
 ],
 "env": {
 "FIRECRAWL_API_KEY": "YOUR_API_KEY_HERE"
 }
 }
 }
} Copy configuration ↗

### Compatibility

An MCP client supporting stdio, Streamable HTTP. · Node.js for the local adapter; a Firecrawl account for authenticated jobs.

### Implementation

TypeScript

### Transports

stdio · Streamable HTTP

### Web extraction and automation / Configured credentials or local access

 - Sends URLs and job inputs to Firecrawl; enabled tools crawl sites and interact with remote pages.
- Authenticates with FIRECRAWL_API_KEY when run locally. The hosted endpoint takes OAuth through /v2/mcp-oauth or the API key in an Authorization: Bearer header; keyless hosted use covers scrape, search and parse only.

### Cost model

Firecrawl service credits/account usage determine cost; limited keyless modes have separate restrictions.

04 / FOLLOW THE EVIDENCE

## The source trail.

Our notes are separate from the original resource. Check upstream before adopting a new version.

 - [Official documentation ↗](https://github.com/firecrawl/firecrawl-mcp-server/blob/1554b675ebb697e2d630d3164bbe7601bb8cdcb0/README.md) Checked 2026-10-13 https://github.com/firecrawl/firecrawl-mcp-server/blob/1554b675ebb697e2d630d3164bbe7601bb8cdcb0/README.md Supports: summary, whySelected, bestFor, limitations, install, access, compatibility, transports, configExample, review
- [Applicable source licence ↗](https://github.com/firecrawl/firecrawl-mcp-server/blob/1554b675ebb697e2d630d3164bbe7601bb8cdcb0/LICENSE) Checked 2026-10-13 https://github.com/firecrawl/firecrawl-mcp-server/blob/1554b675ebb697e2d630d3164bbe7601bb8cdcb0/LICENSE Supports: license
- [Implementation: package.json ↗](https://github.com/firecrawl/firecrawl-mcp-server/blob/1554b675ebb697e2d630d3164bbe7601bb8cdcb0/package.json) Checked 2026-10-13 https://github.com/firecrawl/firecrawl-mcp-server/blob/1554b675ebb697e2d630d3164bbe7601bb8cdcb0/package.json Supports: language, review
- [Implementation: src/index.ts ↗](https://github.com/firecrawl/firecrawl-mcp-server/blob/1554b675ebb697e2d630d3164bbe7601bb8cdcb0/src/index.ts) Checked 2026-10-13 https://github.com/firecrawl/firecrawl-mcp-server/blob/1554b675ebb697e2d630d3164bbe7601bb8cdcb0/src/index.ts Supports: language, review, access
- [Implementation: src/service/index.ts ↗](https://github.com/firecrawl/firecrawl-mcp-server/blob/1554b675ebb697e2d630d3164bbe7601bb8cdcb0/src/service/index.ts) Checked 2026-10-13 https://github.com/firecrawl/firecrawl-mcp-server/blob/1554b675ebb697e2d630d3164bbe7601bb8cdcb0/src/service/index.ts Supports: review, Deprecated credential-in-path route
- [GitHub repository metadata ↗](https://api.github.com/repos/firecrawl/firecrawl-mcp-server) Checked 2026-10-13 https://api.github.com/repos/firecrawl/firecrawl-mcp-server Supports: limitations, review, Repository renamed from mendableai/firecrawl-mcp-server

KEEP COMPARING

## Other approaches to consider.

Related by category or shared topics. These are alternatives to inspect, not a measured quality order.

 [### Apify MCP Server ↗ Searches Apify Store and runs Apify Actors (scrapers, crawlers and automation tools) from an MCP client, through Apify's hosted endpoint or a local stdio package.](/mcp-servers/apify/)[### Bright Data MCP Server ↗ Search-engine results, page scraping, pre-built structured-data scrapers and a remote browser through Bright Data's proxy infrastructure, from a hosted URL or a local npx package.](/mcp-servers/brightdata/)[### Brave Search MCP ↗ Provides web, news, image, video and place search using the Brave Search API.](/mcp-servers/brave/)

 [AI Tools ↗](/tools/)[Skills ↗](/skills/)[Agents ↗](/agents/)[MCP Servers ↗](/mcp-servers/)[Workflows ↗](/workflows/)

## Continue your investigation

 - [Inspect resource evidence](/tools/)
- [Choose a host](/agents/)
- [Inspect reusable workflows](/skills/)
- [Use Undominated data](/api/)
