Open source · ★ 191,296 GitHub stars (as of 2026-10-14)
firecrawl
Firecrawl is an API that searches and collects web pages and turns them into data that LLMs can easily use.
Overview
Firecrawl is a web data API that lets AI agents and apps find, read and interact with web data. It is open source and is also available as a hosted service.
Built around Search, Scrape and Interact, it also offers Crawl, Map, Batch Scrape and Agent features, and returns results as Markdown, structured JSON, screenshots and more. It handles proxy rotation, rate limits and JavaScript-heavy pages for you, and also parses PDFs and DOCX files posted on the web. It can be connected directly to AI agents through MCP or a CLI.
It suits developers who need to feed up-to-date web information into RAG or agents.
What it is used for
- Search the web and get the full content of the result pages.
- Convert a URL into Markdown or structured JSON to use as LLM input.
- Crawl an entire website or find a site's list of URLs.
- Click, scroll and type on a page, then extract its content.
- Add web search and collection tools to AI agents through MCP.
Pricing and license
Open source; the hosted service can be started for free (check the official site for pricing)
Getting started
- Sign up at firecrawl.dev and get an API key.
- Check how it works in the playground first.
- Install the Python SDK with pip install firecrawl-py and call scrape or search with your API key.
- To connect it to an agent, run npx -y firecrawl-cli@latest init --all --browser.
Links
Related terms and companies
Sources
Compiled from public sources (official pages, READMEs). Pricing and features change, so check the official site.