Firecrawl
ServiceWeb scraping API for AI
Use it when your AI agent needs clean web data from JS-heavy pages without managing proxies
Firecrawl turns websites and web searches into LLM-ready data through an API, SDKs and an MCP server.
Added to the shelf
From the makersREADME, read 15 Oct
Firecrawl supplies web data to AI agents: it searches the web and returns full page content, scrapes any URL into markdown, JSON, HTML or screenshots, crawls whole sites, maps a site's URLs, and runs an Agent that gathers data from a plain-language prompt. It handles JS-heavy pages, proxies and rate limits so you do not have to. It is open source and also offered as a hosted service with an API key.
Install
npx -y firecrawl-cli@latest init --all --browserCommands
firecrawl search "firecrawl" --limit 5Search the web and return page content
firecrawl scrape https://firecrawl.devScrape a page to LLM-ready markdown
app.crawl(url) via POST https://api.firecrawl.dev/v2/crawl