FireCrawl MCP Server
@MCP-Mirror
Mirror of
Overview
What is FireCrawl MCP Server?
This server integrates the FireCrawl API into the Model Context Protocol (MCP), enabling advanced web scraping of JavaScript-heavy sites with rate‑limiting, multiple output formats, and batch processing.
How to use FireCrawl MCP Server?
Install via npm install mcp-server-firecrawl, obtain a FireCrawl API key, then add the server to your claude_desktop_config.json with the key set in the FIRE_CRAWL_API_KEY environment variable.
Key features of FireCrawl MCP Server
- JavaScript rendering for dynamic content extraction
- Supports mobile and desktop viewports
- Built‑in smart rate limiting with cooldown
- Output HTML, Markdown, screenshots, and raw text
- Batch scraping of multiple URLs in a single job
- Include or exclude specific HTML tags
Use cases of FireCrawl MCP Server
- Scraping content from single‑page JavaScript applications
- Collecting data across many URLs in one batch operation
- Extracting only the main content or filtering by tags
- Capturing full‑page screenshots for visual verification
FAQ from FireCrawl MCP Server
What output formats does FireCrawl MCP Server support?
It supports markdown, html, rawHtml, screenshot, links, screenshot@fullPage, and extract.
What are the rate limits for the free tier?
Free tier allows 3 requests per minute with a 25‑second cooldown after hitting the limit; higher limits are available on paid plans.
How do I obtain a FireCrawl API key?
Sign up at firecrawl.dev, generate an API key from the dashboard, and set it in the FIRE_CRAWL_API_KEY environment variable.
Can the server scrape multiple URLs at once?
Yes, use the fire_crawl_batch tool to initiate a batch job for multiple URLs, and fire_crawl_status to check its progress.
What authentication does the server require?
It uses an API key passed via the FIRE_CRAWL_API_KEY environment variable. No other authentication is needed.
