@mendableai/firecrawl Mcp Server

What is @mendableai/firecrawl Mcp Server?

@mendableai/firecrawl Mcp Server is an MCP (Model Context Protocol) server that connects any MCP-compatible client to the Firecrawl web data API. It allows AI agents to search, scrape, and interact with the web at scale, converting web content into clean Markdown or structured JSON. The server is intended for developers building AI agents that need real-time web data.

How to use @mendableai/firecrawl Mcp Server?

Configure your MCP client with the following JSON, replacing fc-YOUR_API_KEY with your Firecrawl API key (obtained from firecrawl.dev). The server runs via npx:

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

After adding the configuration, restart your agent. The server exposes tools for search, scrape, crawl, and more.

Key features of @mendableai/firecrawl Mcp Server

• Search the web and get full page content.
• Scrape any URL to markdown, HTML, or structured JSON.
• Crawl entire websites with a single request.
• Interact with pages (click, scroll, type) via AI prompts.
• Agent endpoint for autonomous data gathering.
• Supports media parsing from PDFs, DOCX, and more.
• LLM‑ready output with clean markdown and structured data.
• Handles proxies, rate limits, and JS‑blocked content automatically.

Use cases of @mendableai/firecrawl Mcp Server

• AI agents that need to search and retrieve up‑to‑date web information.
• Extracting pricing, documentation, or product data from multiple websites.
• Automating web research tasks by describing what you need in natural language.
• Crawling a domain to collect all its content for training or indexing.
• Presenting structured data from web pages to LLMs for further processing.

FAQ from @mendableai/firecrawl Mcp Server

What do I need to use this MCP server?

You need a Firecrawl API key, which you can get by signing up at firecrawl.dev. The server runs via npx and requires Node.js installed.

What data does the server access?

The server accesses the Firecrawl API, which scrapes or searches publicly available web pages. No private data is accessed unless you explicitly provide URLs.

Does the server support structured output?

Yes. You can use a schema with the Agent endpoint to get structured JSON data, for example, a list of founders with names and roles.

What are the runtime requirements?

The server is a Node.js package (firecrawl-mcp) run via npx. It requires an internet connection to communicate with the Firecrawl API.

How does the MCP server handle authentication?

Authentication is done via the FIRECRAWL_API_KEY environment variable set in the MCP client configuration. The API key is passed to the Firecrawl API on each request.