← Back to tools

firecrawl

Scrape, crawl, search, and extract web data as LLM-ready content

Homepage Repository
Network Data & JSON Development linuxmacoswindows TypeScript MIT

Description

Firecrawl CLI is a command-line interface for scraping, crawling, searching, and extracting data from websites. Transforms web content into LLM-ready formats including Markdown, HTML, JSON, screenshots, and summaries. Features an AI agent mode for autonomous data extraction using natural language prompts. Integrates as an MCP server and agent skill for AI coding assistants.

Install

npmnpm install -g firecrawl-cli
npxnpx -y firecrawl-cli@latest init --all --browser

AI Summary

CLI for scraping, crawling, searching, and extracting web data into LLM-ready formats — supports AI agent mode, MCP integration, and multiple output formats.

Capabilities

  • + Scrape individual webpages into Markdown, HTML, JSON, or screenshots
  • + Crawl entire domains systematically
  • + Web search with optional content scraping
  • + Map and discover all URLs on a website
  • + AI agent mode for autonomous natural language extraction
  • + JavaScript rendering with wait-for options
  • + Batch processing of multiple URLs
  • + MCP server integration for AI coding assistants
  • + Include/exclude tag filtering

Use When

  • → You need to scrape web content for LLM consumption
  • → You want to crawl a website and extract structured data
  • → You need web search results with full page content
  • → You want AI-powered autonomous data extraction

Avoid When

  • x You need to scrape without an API key or internet access
  • x You only need simple HTTP requests (use curl)

Usage Patterns

Scrape a page as Markdown

firecrawl scrape https://example.com -f markdown

Extract page content as clean Markdown

Search the web

firecrawl search "best CLI tools 2026" --scrape

Search and optionally scrape the result pages

Map a site

firecrawl map https://example.com

Discover all URLs on a website

AI agent extraction

firecrawl agent "find all pricing plans on this site" --url https://example.com

Autonomous AI-powered data extraction

Input / Output

stdout: Extracted content in Markdown, HTML, JSON, or other formats

Typical Pipelines

firecrawl scrape URL -f markdown | llm summarize
firecrawl map URL | xargs -I{} firecrawl scrape {} -f json

Related Tools

slacrawl complementary

Slack-specific data extraction to SQLite
View AGENTS.md for firecrawl