ToolPicker
Firecrawl

Firecrawl

76/100

Turn any site into clean, LLM-ready markdown with one API call.

firecrawl.dev

// scorecard · 76/100

verified 1mo ago
methodology →
Agent-readiness26/35
Agent registration (auth.md)9/9
Public API5/5
OpenAPI spec0/9
MCP server9/9
llms.txt3/3
Reliability & performance12/15
Actively maintained8/8
Production-ready4/4
Updated recently
Pricing10/13
Itemized public pricing6/6
Free tier / trial4/4
Free-tier generosity0/3
Docs & DX10/12
Docs quality8/8
Public API2/2
Quickstart / examples
Security & compliance8/12
SOC 26/6
ISO 270010/2
GDPR0/2
Security / trust page2/2
Openness10/10
Open source6/6
Open / un-gated API4/4
Company maturity0/3
Company size
Years operating

// rankings · cost per standard workload

#3 in Web Search$333/ 100k queries

Search 2cr/10 results → 100k=200k cr → Growth $333/mo

firecrawl.dev/pricing
#12 in Scraping$83/ 100k pages

1 page = 1 credit (basic scrape). Standard plan $83/mo = 100k credits. JS-render/extract cost more.

// overview

Firecrawl is a developer-first web data API for AI agents that searches, scrapes, crawls, maps, parses, and interacts with web pages to return LLM-ready Markdown, JSON, screenshots, and structured data.

Best for: Developers, AI teams, GTM/revops builders, data teams, and agent/RAG product teams that need reliable, structured live-web data without maintaining scraping infrastructure.

// pricing

Free tier · generosity 0/5
Includes: Scrape 1000 pages · 2 concurrent requests · Low rate limits
Limits: 1000 credits/month
Free0
  • 1000 credits/month
Hobby16
  • 5000 pages
Standard83
  • 100000 pages
Growth333
  • 500000 pages
Scale599
  • 1000000 pages
EnterpriseCustom
  • unlimited pages

// security & compliance

SOC 2 ISO 27001 GDPR HIPAA

// features

  • Search the web and return full page content from results
  • Scrape URLs into Markdown, HTML, structured JSON, screenshots, and metadata
  • Crawl entire sites recursively and gather page content
  • Map/discover URLs on a website
  • Batch scrape thousands of URLs asynchronously
  • Interact with pages by clicking, filling forms, navigating, or extracting dynamic content
  • Parse PDFs, documents, spreadsheets, and other hosted files into LLM-ready data
  • JSON mode and natural-language structured extraction
  • Monitoring/change tracking and webhooks
  • Official SDKs for Python, Node/TypeScript, Go, Java, Ruby, Rust, .NET, PHP, Elixir plus CLI and MCP server

pros

  • Broad web-data API surface built specifically for AI agents and LLM/RAG workflows
  • Hosted service abstracts away proxies, JavaScript rendering, anti-bot handling, orchestration, and rate limits
  • Generous free tier and clear credit-based plans for common use cases
  • Open-source with self-hosting path, SDKs, CLI, MCP, and many AI-framework integrations

cons

  • Credit pricing can become expensive or harder to predict at high volume or with advanced features
  • No simple pay-per-use plan; users must choose monthly plans or custom Scale/Enterprise
  • Credits generally do not roll over between periods
  • Some dynamic/interactive or anti-bot-heavy sites may still require testing and tuning

sources