ai-search
Firecrawl
AI-native web scraping + crawling API — LLM-ready data extraction
Turns any website into clean LLM-ready markdown/JSON. Handles JS, anti-bot, PDFs, screenshots. Scrape + Crawl + Map + Extract (with AI) + Batch endpoints. OSS self-host option.
Pricing
| Tier | Price | Notes |
|---|---|---|
| Free | Free | 500 credits one-time. 2 concurrent requests. For exploration. |
| Self-Hosted (OSS) | Free | AGPL-3.0 licensed. Run on your own infra. |
| Hobby | $19/mo | $19/month ($16 annual). 3,000 credits/month. 5 concurrent. |
| Standard | $99/mo | $99/month ($83 annual). 100,000 credits. 50 concurrent. Most popular. |
| Growth | $399/mo | $399/month ($333 annual). 500,000 credits. 100 concurrent. |
| Enterprise | Custom | Custom. Dedicated support + SLA, bulk discounts, zero-data retention, SSO, advanced security. |
Limits
| Tier | Metric | Value | Notes |
|---|---|---|---|
| — | advanced features cost | Some features (JS render, stealth, ai-extract) cost additional credits | Advanced features |
| — | ai extract | Define schema → Firecrawl runs LLM extraction for you | AI Extract |
| — | anti bot | Stealth mode + residential proxies on Standard+ | Anti-detection |
| — | credit per page | 1 credit per webpage or PDF page | Base credit cost |
| — | credit rollover | Plan credits do NOT roll over. Auto-recharge packs DO. | Rollover policy |
| — | javascript rendering | Headless browser rendering for SPAs + JS sites | JS support |
| — | max crawl pages standard | 100k pages/run cap on Standard | Crawl scale |
| — | output formats | markdown, HTML, rawHtml, structured (JSON), screenshot, screenshot@fullPage, links | Output formats |
Features
- Batch Scrape — Scrape many URLs in parallel.
- Crawl — Crawl an entire site following links. Returns all pages. · docs
- Custom Headers/Cookies — Scrape behind login via headers.
- Deep Research (beta) — AI agent that autonomously researches a query across many sites.
- Extract (AI) — Structured extraction with a JSON schema → Firecrawl runs LLM for you. · docs
- JavaScript Rendering — Headless browser handles SPAs, React/Vue sites.
- LLM-Ready Markdown — Cleaned + normalized Markdown optimized for LLM ingestion.
- Map — Fast — returns all URLs on a site without fetching content. · docs
- MCP Server — Official MCP for Claude / Cursor / Agents.
- Open Source (AGPL) — Self-host the entire stack via Docker.
- PDF Parsing — Extract text from PDFs.
- Residential Proxies — Bypass anti-bot + geoblocks (paid).
- Scrape — Single URL → clean markdown/JSON/HTML. · docs
- Screenshots — Visual screenshot alongside content.
- Search (beta) — Google-like search API returning results + content.
- Structured Output — Extract to JSON schema via LLM.
Developer interfaces
| Slug | Name | Kind | Version |
|---|---|---|---|
| cli | Firecrawl CLI | cli | 0.x |
| sdk-go | firecrawl-go | sdk | 0.x |
| sdk-node | firecrawl-js | sdk | 1.x |
| mcp | Firecrawl MCP Server | mcp | — |
| sdk-python | firecrawl-py | sdk | 1.x |
| rest-api | Firecrawl REST API | rest | v1 |
Compare Firecrawl with
ai-api
Firecrawl vs Anthropic API
Side-by-side breakdown.
ai-api
Firecrawl vs AssemblyAI
Side-by-side breakdown.
ai-api
Firecrawl vs Deepgram
Side-by-side breakdown.
ai-api
Firecrawl vs ElevenLabs
Side-by-side breakdown.
ai-api
Firecrawl vs Google Gemini API
Side-by-side breakdown.
ai-api
Firecrawl vs Groq
Side-by-side breakdown.
ai-api
Firecrawl vs OpenAI API
Side-by-side breakdown.
ai-api
Firecrawl vs Replicate
Side-by-side breakdown.
Staxly is an independent catalog of developer platforms. Outbound links to Firecrawl are plain references to their official pages. Pricing is verified at publication time — reconfirm on the vendor site before buying.