Crawl4AI Plus AdvancedInstall

Summary

Stream crawl one or more URLs by sending them to the Crawl4AI streaming endpoint, receiving full crawl results with advanced browser session options, crawl settings, and output filtering for precise content extraction.

Inputs

• URLs (required) — One or more URLs to crawl using the streaming endpoint, entered one per line or comma-separated. Each URL generates a separate output item.
• Browser & Session — Options to configure the browser profile, engine, session state, headers, cookies, JavaScript execution, stealth mode, proxy, viewport, timeout, and related browser behaviors to control how pages are loaded and rendered.
• Crawl Settings — Parameters controlling crawl behavior such as anti-bot measures, resource blocking (ads, CSS), caching mode, JavaScript execution, retry attempts, waiting conditions before extracting content, and content scope selectors.
• Output & Filtering — Controls output format and content filtering strategies including content filters (BM25, LLM, pruning, or none), table extraction method, markdown output type, inclusion of media, links, HTML, screenshots, PDFs, and verbosity/debug options.

Output shape

a list of items, one per input URL, each containing the full crawl result including extracted content, metadata, and filtered data according to user-specified options

The node streams the full crawl results from multiple URLs using the streaming endpoint and buffers them before returning. The output respects all specified crawl and output filtering configurations. Errors in JSON parsing of streamed data are surfaced on the last output item. Content filtering and table extraction can be optionally applied requiring relevant credentials. Output may include base64-encoded screenshots or PDFs if requested.

Links

• n8n community nodes installation guide