How to Use Firecrawl MCP for Web Scraping

What is the Firecrawl MCP?

Firecrawl is a web scraping and crawling API that handles JavaScript rendering, bot detection bypass, and content extraction automatically. The Firecrawl MCP brings these capabilities into Claude Code, letting Claude scrape individual pages, crawl entire sites, and extract structured data — without you managing browser automation or proxies.

Installation

mcpizy install firecrawl

Create an account at firecrawl.dev and copy your API key. Paste it during MCPizy setup.

Key Capabilities

• Scrape single pages — extract clean Markdown, HTML, or structured data from any URL, including JavaScript-rendered content.
• Crawl entire websites — recursively follow links and scrape every page up to a configurable depth.
• Map site structure — discover all URLs on a domain before deciding what to scrape.
• Extract structured data — define a JSON schema and Firecrawl returns structured objects from unstructured web content.
• Search and scrape — combine a web search with immediate scraping of the top results.

Example Usage

Extract all pricing information from a competitor's pricing page:

// Claude will:
// 1. Call firecrawl.scrape("https://competitor.com/pricing")
// 2. Receive clean Markdown with all pricing tiers
// 3. Parse the tiers, features, and prices
// 4. Format a comparison table against your own pricing

Tips and Best Practices

• Use structured extraction with a defined schema rather than asking Claude to parse raw Markdown — it is more reliable.
• Use site mapping before crawling to estimate scope and avoid unexpected costs.
• Combine with Playwright MCP when you need to interact with a page (login, fill forms) before scraping it.
• Respect robots.txt and site terms of service — Firecrawl makes it easy to scrape aggressively, so use it responsibly.

View on the MCPizy directory