Check the links on a page
Paste a page address. We test all its external links and a sample of its internal ones, and tell you which are followed, nofollow, sponsored or ugc.
Checking the links
Loading the page…
We call every external link and a sample of the internal ones, several at a time. Allow a few seconds, twenty at most.
This is taking longer than usual: a server is slow to answer. We give up on it after five seconds.
How to read this result
The right-hand column shows three verdicts, not two. The distinction matters: most of the “broken link” reports going around flag links that work perfectly well.
Dead
The server answered with an error, usually 404 (the page does not exist) or 410 (it was deliberately removed). This is the only case where there is really something to do: fix the address, replace it, or drop the link.
Blocks bots
The server answered 403, 429 or 503. That tells you nothing about the page: it tells you a firewall pushed us aside. Plenty of large sites turn away automated requests on principle, and they are often the ones most worth citing. Removing those links means deleting links that open perfectly for your readers.
No answer
Nothing came back within the five seconds allowed. A slow server is not a dead server, and a three-second network hiccup is not a vanished page. Re-check these later before touching anything.
Internal links, external links
A broken link on your own site can be fixed right now: it is yours, you have the keys. A broken link on someone else's cannot, the page at the far end is gone. It is still worth knowing, because an article riddled with dead links ages badly, for your readers and for Google alike.
What this tool does not do
- It looks at one page only: all its external links, and ten of its internal ones. A whole site needs a crawl, not twenty runs of this page.
- It only sees links present in the served HTML. Anything a script adds afterwards escapes it.
- It ranks nothing. A broken link in a footer and one in the middle of an article do not carry the same weight, and that call is yours.
Frequently asked questions
A link that works is marked “blocks bots”. Why?
Because the site guards itself against robots, and we are one. Its firewall answers 403 to any automated request, whoever sends it. The link opens normally in a browser, and we would rather say so than wrongly condemn it.
How long does the check take?
A few seconds, twenty at most. Links are called several at a time, and each one gets five seconds to answer; after that it is marked “no answer”.
Why every external link, and only ten internal ones?
Because an external link dies without you knowing: it is the site on the other end that removes a page. We test every one of those. Internal links are checked across the whole site, page after page: that is a crawl's job, not a free page's. Every link tested leaves our servers for a third-party site, and this cap also protects the sites you test.
Are nofollow links tested?
Yes. The rel attribute tells search engines not to follow the link; it says nothing about the state of the page at the far end. A nofollow link to a dead page is still a dead link for your readers.
What is the difference between nofollow, sponsored and ugc?
All three ask the search engine not to pass credit to the target site. Sponsored marks a paid link, ugc a link posted by a user (comment, forum), nofollow everything else. Since 2019 Google reads them as hints rather than orders. A paid link carrying neither sponsored nor nofollow breaks its rules, and this table is how you spot one.
Going further
Find broken links by hand, without a paid tool Four free ways to find the broken links on a site, what each one misses, and the point where doing it by hand stops being reasonable.The other tools
Can my page be indexed?
Noindex, robots.txt, canonical, HTTP status: what stops Google and Bing from indexing a page, engine by engine.
Test the robots.txt
Whether the site robots.txt allows this address to be crawled.
Robots.txt generator
A robots.txt written for your site, AI crawlers included, tested before it goes live.
One page by hand, or the whole site on its own
Expansel walks your pages, tests the links on each one and hands you the list of what is broken, with the page where each dead link was found. One page by hand here, the whole site in a single crawl there.