Map a small route
Request up to five same-site pages and extract links from the HTML responses.
A focused website link check
Find broken links without crawling forever.
Sample a handful of public pages, trace their links and see where visitors may meet a clear 404, an avoidable redirect or a response that needs a closer look.
Check a websiteChoose the starting point
Where should the sample begin?
Use a homepage or an important section page. The checker follows same-site links from there and reports the pages it actually reached.
This is a sample: up to five same-site pages and thirty links. It checks public responses and follows the crawl rules published in robots.txt.
How the check behaves
Careful requests, useful distinctions.
The checker begins with robots.txt, stays within its page and link limits, and validates every destination before connecting. Link probes use HEAD where supported and fall back to a small GET only when a server rejects HEAD.
Test a balanced sample
Check up to thirty internal and external HTTP links with limited parallel requests.
Separate fact from uncertainty
Reserve “broken” for clear 404 or 410 responses and explain everything else in its own category.
Broken link questions
Read the result in context.
Does this check every page on a website?
No. It samples up to five same-site pages and thirty discovered links. The coverage panel shows exactly what this run included.
Why is a link marked “could not verify”?
A timeout, access rule or server error does not prove that normal visitors receive a broken page. Open an important link in a browser and check again later before removing it.
Does the checker respect robots.txt?
Yes. It reads each origin's published rules before requesting a page and skips URLs that do not permit this crawler.
Why can my browser show something different?
Websites can vary responses by session, location, device, request method or time. The report records one anonymous observation, not every possible visitor experience.