| name | firecrawl |
| description | Web scraping, site crawling, search, structured data extraction, and AI-powered research with Firecrawl CLI. Use when you need full page content as markdown, JS-rendered pages, anti-bot bypass, crawling entire documentation sites, extracting structured data with schemas, or deep web research. Prefer WebFetch for quick questions about a known URL. Prefer WebSearch for finding links without full content. |
Firecrawl
When to Use What
| Need | Tool | Why |
|---|
| Quick answer from a known URL | WebFetch | Free, fast, no API key needed |
| Find links/pages on a topic | WebSearch | Free, fast, returns snippets |
| Full page content as clean markdown | Firecrawl scrape | Handles JS rendering, anti-bot, better extraction |
| Scrape JS-heavy SPAs or paywalled sites | Firecrawl scrape | --wait-for, proxy support |
| Discover all URLs on a site | Firecrawl map | Sitemap + link discovery |
| Crawl entire docs/blog site | Firecrawl crawl | Recursive, depth control, path filtering |
| Search + get full page content | Firecrawl search | --scrape flag fetches each result |
| Extract structured data with schema | Firecrawl extract | LLM-powered, typed JSON output |
| Open-ended web research | Firecrawl agent | AI browses autonomously, multi-step reasoning |
Commands
| Command | Usage |
|---|
/firecrawl:scrape | <url> [--only-main-content] [--format markdown,links] |
/firecrawl:search | <query> [--limit N] [--scrape] [--tbs qdr:d] |
/firecrawl:map | <url> [--search "filter"] [--limit N] |
/firecrawl:crawl | <url> --wait [--limit N] [--max-depth N] [--include-paths "/api/*"] |
/firecrawl:extract | <url> --prompt "..." [--schema '{...}'] |
/firecrawl:agent | "<prompt>" [--urls <urls>] [--wait] [--model spark-1-pro] |
Prerequisites
- API key:
export FIRECRAWL_API_KEY=fc-your-key or npx firecrawl login
- CLI:
npm install -g firecrawl-cli (or use via npx firecrawl)
- Get key from: https://firecrawl.dev
Credit Costs
| Operation | Cost |
|---|
| Scrape | 1 credit/page |
| Map | 1 credit/call |
| Search | 1 credit/result |
| Crawl | 1 credit/page crawled |
| Extract/Agent | Varies (typically 5-50 credits) |
Check https://firecrawl.dev/pricing for current rates.
Detailed References
For full flags and examples per command, see:
references/scrape-reference.md
references/search-reference.md
references/map-reference.md
references/crawl-reference.md
references/extract-reference.md
references/agent-reference.md