Scrape
Turn any public web page into clean Markdown for LLMs, with its metadata, links, images, structured data and AI-use signals.
Scrape reads one page as plain HTML, the way AI crawlers see it, and returns it as Markdown (the whole page as a person reads it, or only the article with scope=main) together with everything else on the page that a model or an SEO check needs. One call for one page, or a job for up to 10,000.
Pages
Scrape a pageTurn a web page into clean Markdown for LLMs, with its metadata, headings, links, images, structured data, language and AI-use signals.Start a Scrape jobScrape up to 10,000 URLs in one background job; each item takes the same parameters as a single call.Get a Scrape jobCheck a Scrape job: its status, how many items are done and failed, and when its results expire.Get Scrape job resultsRead a Scrape job's items as each one finishes, with its full answer or error, and page through them with a cursor.Cancel a Scrape jobStop a Scrape job. Items already finished stay readable for 24 hours; items not started are dropped.