Skip to main content
The Scrape category fetches a page and returns content your agent can use directly. It is small — four tools from three providers — so the choice is mostly about how hard the target page is to reach.

​
Explore

URL to markdown

Any page or PDF as clean, LLM-ready text.

Anti-bot pages

Challenge-gated pages returned as raw HTML.

​
Providers

​
Choosing a tool

  • Start with Firecrawl or Jina Reader for ordinary pages.
  • Fall back to BrightData Web Unlocker only when a page is protected by anti-bot challenges — it returns the raw response, not cleaned text.
  • Need many URLs or a whole site? Tavily/post_extract, Tavily/post_crawl and Tavily/post_map live in the Search category.
Building with code? Developers and enterprise integrations can call this category through Scenario Methods: fixed REST contracts with a stable schema, a fixed price and automatic provider failover. The methods are listed by platform in the API Reference sidebar.

​
Browse the catalog

Parameters and per-call cost differ between the four tools — check describe_tool before executing. execute_tool calls exactly the tool you name; switching to another scraper on failure is your call, not automatic.