Jina Reader returns empty content for a normal page (bot-prevention DOM swap)

Jina Reader returns empty content for a normal page (bot-prevention DOM swap)

What's going on

Jina AI Reader returns empty Markdown content for an ordinary page (the Government of Canada gender-based violence glossary), even though the page is simple HTML with no obvious anti-scraping protection. Interestingly, pageshot mode renders the full page, so the browser itself loads it fine.

The verified fix

The maintainer found the page's CDN bot-prevention layer ("edgesuite") was swapping the DOM for warning content before Reader captured it, so Reader faithfully extracted the wrong DOM. The fix shipped on the Reader backend; the original reporter later confirmed the exact same link now returns the full page as Markdown. If you hit empty output on a page the browser clearly renders, retry it, since the extraction-side fix applies to the hosted endpoint.

Source: https://github.com/jina-ai/reader/issues/105

How to use this

Take the question above and check if it matches what you're seeing. If it does, work through the verified fix step by step. Start with the cause described first, because that's what tells you the fix applies to your setup, then apply the changes in the order given. Verify by re-running whatever failed before, and expect the same behavior the thread author reported.

Someone asked: Jina AI Reader returns empty Markdown content for an ordinary page (the Government of Canada gender-based violence glossary), even though the page is simple HTML with no obvious anti-scraping protection. Interestingly, pageshot mode renders the full page, so the browser itself loads it fine. The verified answer: The maintainer found the page's CDN bot-prevention layer ("edgesuite") was swapping the DOM for warning content before Reader captured it, so Reader faithfully extracted the wrong DOM. The fix shipped on the Reader backend; the original reporter later confirmed the exact same link now returns the full page as Markdown. If you hit empty output on a page the browser clearly renders, retry it, since the extraction-side fix applies to the hosted endpoint.

Source: https://github.com/jina-ai/reader/issues/105 Discussion and confirmation live at https://vectle.com/threads/thr_if03Y9IvC3ybqqKTKlSf6Q