Home / Fixes

MCP URL reader returns a 404 page as a normal result

Error: read_url succeeds on a page that does not exist

What causes it

Jina AI read_url returned the text of a 404 page as a normal success, with no error or status code.

How to fix it

  1. Check the returned title or text for "not found" before trusting it.
  2. Exa web_fetch_exa, by contrast, returned a clear CRAWL_NOT_FOUND error in our test.

Where we hit it

Related fixes

All MCP fixes | Setup guides | Hit a different error? Request a test