Paste a URL and get the page's actual content back, headings, paragraphs and links, with the navigation, cookie banners, ads and footers stripped out. It renders JavaScript, so single-page apps come through too. Useful for feeding a page into an AI prompt, archiving a reference, or pulling every link off a page.
Public pages only. Check the site's terms and robots.txt before scraping at scale.
Extracting a page you can already read in a browser is normally fine. Most of the trouble comes from what happens next.
Copy-paste brings the navigation, sidebars, cookie notices and footer with it, and loses heading structure. This returns just the article, as markdown, with headings and links intact, which is what you want when pasting into an AI prompt or a document.
Usually. The page is rendered before extraction, so most single-page apps come through. If the reader is busy the tool falls back to a direct fetch and tells you so in the output, because a direct fetch can miss JavaScript-rendered content.
No. It only fetches pages that are publicly reachable without credentials, exactly as any visitor would see them.
The reader service allows a shared 20 requests a minute. If it is busy you will get a direct fetch instead, with a note in the output. Wait a minute and re-run for the full rendered version.
Reading a public page is generally fine; republishing it is a different question. Check the site's terms and robots.txt, don't hammer a server, and don't republish someone's content as your own. The do's and don'ts below cover the practical rules.