Dead links waste crawl budget, frustrate readers and weaken trust. Paste a URL — we pull the links on the page and check each one's status (up to 25 per run).
A dead hreflang alternate does not break one link — it breaks the
cluster. If any language version in a set returns 404, Google discards the
whole set, and the versions that do work lose their language signals
with it. One dead URL, several pages affected, and nothing on any of them
looks wrong.
We learned this by checking our own site: 414 pages, scanned on
every commit over the built HTML — internal <a href> and every
hreflang alternate. Today: 0 broken. The number is boring;
the discipline is the point. Both classes are invisible while you develop —
the build passes, the page opens, and only a click or a crawler finds the hole.
What this measurement is and is not. One site, one build pipeline, since 2026-07. It says nothing about how many broken links a typical site has; it says that a per-commit scan of the rendered output keeps that number at zero, and that hreflang belongs in the scan.
A dead link wastes the two things you spent most to earn: a visitor's patience and a crawler's budget. Search engines allocate a finite number of fetches per site — every one spent on a 404 is one not spent on a page you want indexed.
Broken internal links are the worse half: they also break the path authority flows along, so pages behind the dead link get less of it than you intended.
Links assembled by a framework, a CMS plugin or a template partial do not exist as such in your repository. A scan of the built HTML is the only one that sees what Googlebot sees — which is why our own check runs over the build output, never over source files.
The practical rule mirrors the one for structured data: whatever you validate must be the same bytes the crawler gets. Paste the live URL above.
Internal breakage is yours to fix outright: update the URL or redirect the old path. It usually appears after a restructure, when addresses changed but references did not.
External breakage is someone else's site going away. You cannot fix it, so choose: replace the link with a live equivalent, point at an archived copy, or drop it. Leaving it is the only wrong answer — a citation to nothing weakens the passage it was meant to support.
This tool scans links out of a page. Links into your site rot too, and that rot is lost authority you once had. On 9 sites we monitor, one had 51 live backlinks and 64 broken — more dead than alive; the strongest profile carried 235 broken out of 2,985.
Those numbers come from a backlink tracker, not from this checker; the two sides need two tools. Authority scores summarise the live part only — see the domain authority checker for why a profile that is mostly rot looks bigger than it is.
Enter your URL above. The checker crawls the page's links and reports every one that answers with an error, split into internal and external.
Indirectly but measurably. They waste crawl budget, break the internal paths authority flows along, and push visitors away — all three reduce how much of your site gets indexed and ranked.
A broken link returns an error (404, 410, 5xx) and leads nowhere. A redirect returns 301/302 and forwards to a working page. Redirects are fine; long chains of them are not.
Yes — hreflang. If one language alternate in a cluster returns 404, Google discards the whole cluster, so the working versions lose their language signals too. One dead link, several pages affected. We check for exactly this on every build of our own site.
The rendered page. Links assembled by a framework, a CMS plugin or a template partial do not exist as such in your source; a scan of the built HTML is the only one that sees what a crawler sees. That is how we run our own check — over the build output, not the repository.
Different side, same rot. On nine sites we monitor, one had 51 live backlinks and 64 broken ones; the strongest profile carried 235 broken out of nearly 3 000. Broken inbound links are lost authority you once had; this tool scans outbound links on a page, and a backlink tracker shows the inbound side.
A broken link checker scans the links on a page and flags the ones that return a 404 or resolve through a dead redirect instead of a working destination. Dead links waste crawl budget, break the paths crawlers and AI answer engines follow to map your site, and erode the trust that makes a page worth citing. Fixing them keeps link equity flowing and keeps your content reachable to both search bots and the models that quote it.
A broken link is one whose destination returns an error (commonly a 404 Not Found or a 5xx server error) or forwards through a redirect that ends in an error or loop, so the visitor or crawler never reaches a working page.
They waste crawl budget on dead ends, block the link paths crawlers and AI engines use to map your site, and signal low maintenance — all of which lower the odds a page gets indexed, ranked, or quoted.
It scans up to 25 links from the URL you paste — enough for the main in-content links on many pages. For link-heavy pages, run it on your most important pages one at a time.
No — the checker is free and requires no signup; paste a URL and get the results in your browser.
Free tools show you WHAT to fix. The full audit + the autonomous team of 20 AI agents fix it end-to-end — strategy, content, GEO, internal links, publishing and day-2 upkeep.
All free tools →cookies
The Matrix already knows everything about you — cookies are small change by comparison. The choice, as always, is yours: