Firecrawl API - Standard Operating Procedures

Updated Feb 2026

Turn any website into clean, LLM-ready data with Firecrawl's powerful web scraping API.

What is Firecrawl?

Firecrawl is an AI-powered web scraping and data extraction API that handles the complexity of modern web infrastructure. It manages proxies, JavaScript rendering, rate limits, and dynamic content while delivering clean, structured data.

Core Capabilities

Installation

# Python SDK
pip install firecrawl-py

# Node.js SDK
npm install @mendable/firecrawl-js

# CLI (includes browser support)
npx -y firecrawl-cli@latest init --all --browser

# MCP Server (for Claude Code)
claude mcp add firecrawl -e FIRECRAWL_API_KEY=fc-YOUR-KEY -- npx -y firecrawl-mcp

Quick Start

1. Get Your API Key

Sign up and get your API key

2. Make Your First Request

curl -X POST 'https://api.firecrawl.dev/v2/scrape' \
  -H 'Authorization: Bearer fc-YOUR-API-KEY' \
  -H 'Content-Type: application/json' \
  -d '{"url": "https://example.com"}'

3. Get Clean Markdown

{
  "success": true,
  "data": {
    "markdown": "# Example Domain\n\nThis domain is for use in illustrative examples...",
    "metadata": {
      "title": "Example Domain",
      "sourceURL": "https://example.com"
    }
  }
}

Key Features

Scrape

Extract content in multiple formats: markdown, HTML, JSON, screenshots, links, images, brand identity, summaries.

Search

Web search across multiple sources (web, news, images) and categories (GitHub, research, PDF). Optionally scrape results.

Map

Discover all URLs on a website instantly. 1 credit per call regardless of URLs returned.

Crawl

Recursively crawl entire websites. Handles JavaScript, respects robots.txt, supports WebSocket watching and webhooks.

Browser Sandbox

Fully managed, isolated browser environments. Pre-installed Playwright, agent-browser CLI with 40+ commands, CDP access, persistent profiles, and embeddable Live View.

Agent (Research Preview)

AI-powered autonomous deep research. Describe what you need — no URLs required. Uses Spark-1 Mini/Pro models with structured schema output.

Change Tracking

Monitor websites for content changes. Git-diff and JSON field-level comparison with persistent snapshots that never expire.

MCP Server

Connect Firecrawl to Claude, Cursor, VS Code, and other AI tools. 12 tools covering all features.

Choosing the Right Tool

Available SDKs

• Python: pip install firecrawl-py
• Node.js: npm install @mendable/firecrawl-js
• CLI: npm install -g firecrawl-cli
• MCP: npx -y firecrawl-mcp
• cURL: Direct HTTP API access

Documentation Sections

• Features → - Detailed guide for each capability
• SOPs → - Step-by-step procedures for common tasks
• Workflows → - Complete implementation patterns
• Quick Reference → - API endpoints, pricing, troubleshooting

Getting Help

• Full API Reference
• Interactive Playground
• Agent Playground
• Community Discord
• Blog & Updates

Pro Tip

Use the chat widget in the bottom right to ask questions about Firecrawl features, pricing, and best practices!

Cost Overview

Operation	Credits
Scrape	1 per page
Search	2 per 10 results
Map	1 per call
Crawl	1 per page
Browser	2 per minute
Agent (Mini)	Dynamic (~60% cheaper)
Agent (Pro)	Dynamic
JSON Mode	+4 per page
Change Tracking (JSON)	+5 per page
Enhanced Proxy	+4 per page

---

Ready to get started? See SOPs for step-by-step guides!