Hanzo Crawl
Turn any website into clean, structured data for LLMs
Drop-in replacement for Firecrawl, Apify, and Scrapy. Powered by crawl4ai.
Crawl anything, extract everything
From static HTML to JavaScript-heavy SPAs -- get structured, LLM-ready data from any website.
LLM-Ready Output
Converts any webpage into clean Markdown with structured metadata. Ready for ingestion by any LLM, RAG pipeline, or Hanzo Search index.
JavaScript Rendering
Full Playwright browser engine handles SPAs, dynamic content, lazy loading, and infinite scroll. See the page exactly as users do.
Structured Extraction
Extract data using CSS selectors, XPath expressions, or LLM-based schemas. Get typed JSON from any page layout.
Parallel Crawling
Crawl thousands of pages concurrently with configurable rate limiting. Respect robots.txt while maximizing throughput.
Direct Search Integration
Pipe crawled content directly into Hanzo Search and Hanzo Vector. One pipeline from raw web to searchable, embeddable knowledge.
Anti-Detection & Stealth
Built-in stealth mode with proxy rotation, user-agent management, and fingerprint randomization for reliable large-scale crawls.
How it works
Three steps from any website to structured, searchable data.
Point at any URL
Provide a seed URL or sitemap. Hanzo Crawl discovers pages, follows links, and handles pagination. Configure depth limits, URL filters, and concurrency to match your needs.
Extract structured data
Each page is rendered with a full Playwright browser, then converted to clean Markdown. Use CSS selectors, XPath, or LLM-based schemas to extract typed data from any layout.
Index and search
Push extracted content directly into Hanzo Search for full-text and hybrid search, or into Hanzo Vector for semantic embeddings. One pipeline from raw web to queryable knowledge.
Works with any content source
Crawl marketing sites, documentation portals, e-commerce catalogs, forums, and more. REST API for programmatic control.
Pricing
Pay for hosted crawling and rendering. Self-host crawl4ai free forever.
Build
- 50,000 pages/month
- 5 concurrent crawlers
- JavaScript rendering
- Markdown + structured output
- REST API access
Scale
- 500,000 pages/month
- 25 concurrent crawlers
- Priority JS rendering
- LLM-based extraction
- Direct Search + Vector integration
Enterprise
- Unlimited pages
- Unlimited concurrent crawlers
- Dedicated rendering pool
- Custom extraction schemas
- Dedicated support + SLA
crawl4ai is open source (Apache 2.0). Self-host free forever. Pay only for our hosted API, JS rendering, and managed infrastructure.
25% of compute goes back to open source
Every deployment is SBOM-verified. Contributors to Crawl4AI earn a share of compute revenue — transparent, on-chain, and customizable by the community.
Turn the web into structured knowledge
Crawl once, use everywhere.
Clean Markdown, structured data, and direct integration with Search and Vector.