Free link checker
Broken Link Checker
Find broken links, and the links AI can't see.
Siteiz tests every link on a page for 404s, dead sites and redirect detours, then pins each problem on a screenshot of your page. It also flags links added by JavaScript, which AI crawlers like GPTBot never see.
Free · no signup · results in about a minute
How it works
How to check a page for broken links
Four steps, about a minute.
- 01
Enter the page URL
Paste the address of the page you want to check, for example your homepage or a product page. No account needed.
- 02
Siteiz reads the page twice
Once as raw HTML, the way AI crawlers receive it, and once in a real browser, the way people see it. Links that only appear in the browser are flagged as hidden from AI.
- 03
Every link is tested
Each link gets a live request. Redirects are followed hop by hop, and each link gets a plain result: works, page not found, detours, or could not verify.
- 04
Fix what is marked
Start with Needs fixing. Each link shows where it sits on the page and one next step, so you or your developer know exactly what to change.
What it finds
Broken links are half the story
A link can work for people and still be invisible to AI. Siteiz checks both, and tells you where each one is.
Broken links
Pages that return 404 or 410, servers that error, and domains that no longer answer.
Detours
Redirect chains and loops that slow visitors down and make crawlers give up early.
Links hidden from AI
Links added by JavaScript. People can click them, but GPTBot and ClaudeBot never see them.
Exactly where they are
Every problem is pinned on a screenshot of your page and named by area: header, menu, content, footer.
The part other checkers miss
Links that work for people but not for AI crawlers
Most link checkers only ask whether a link returns a working page. There is a second question: can a crawler find the link at all? Menus, "related" widgets and footers built with JavaScript often leave the server HTML empty and add the links later in the browser.
Googlebot renders JavaScript, so those links usually count for classic search. AI crawlers such as GPTBot, ClaudeBot and PerplexityBot read the raw HTML and move on. For them, a JS-only link does not exist, and neither does the path to the page behind it. On JavaScript-heavy sites that can be most of the navigation.
The fix is to render navigation and key internal links on the server, as plain <a href> elements. Read more in Does ChatGPT read JavaScript?
Reading the results
What each link result means
We only call a link broken when the evidence says so. Anything less certain gets its own label.
- Page not foundBroken
- The page does not exist (HTTP 404 or 410). Fix or remove the link.
- The site behind it is failingServer error
- The server failed to answer (HTTP 5xx). It may be temporary, so re-check before removing the link.
- Website does not respondUnreachable
- No response: the domain does not resolve, refused the connection, or timed out.
- Goes round in circlesRedirect loop
- The link redirects in a circle or through too many hops, so it never lands on a page.
- The site rejects the linkClient error
- The server rejected the request (a 4xx code other than 404, 410, 401, 403 or 429).
- Takes several detoursRedirect chain
- Two or more redirects before the final page. It works, but each hop costs crawl time. Link to the final URL.
- Could not verifyBlocked
- The site refused our checker (HTTP 401, 403 or 429). Often bot protection, not proof the page is gone. Open it in a browser to confirm.
- Forwards to a new addressRedirect
- One redirect, then a working page. Fine, though linking to the final URL is cleaner.
- WorksOK
- The link answered with a working page.
- Not checkedSkipped
- Not checked for safety reasons: a private network address or a non-standard port.
- Not checkedNot checked
- Over the free limit or the time budget for one run. Not a verdict on the link.
Learn more
Keep your links healthy
How to find and fix broken links
A practical routine: what to check first, how to fix each status, and how often to re-check.
Does ChatGPT read JavaScript?
Why links and content added by scripts are invisible to AI crawlers.
What is an AI visibility audit?
The wider picture: how AI systems access, read and cite your pages.
FAQ
Broken link checker questions
Is this broken link checker free?
Yes. Checking one page is free, with no account and no email required. Each check covers up to 150 unique links on that page, and you can download the full result as a CSV.
How do I find broken links on my website?
Enter a page URL above and run the check. Start with the pages that share your site's templates, such as the homepage, a category page, a product or service page, and an article. A broken link in the header or footer shows up on every page, so fixing it once fixes it everywhere.
How do I see where a broken link is on the page?
Every problem link gets a number. The same number is pinned on a screenshot of your page, and the list names the area it sits in, such as Header, Main content under a specific heading, or Footer. Click a number to jump between the screenshot and the list.
What is a link hidden from AI?
A link that is missing from the HTML your server sends and only appears after JavaScript runs in a browser. People can click it. AI crawlers such as GPTBot, ClaudeBot and PerplexityBot read the raw HTML and do not run scripts, so they never see it or follow it.
Why are some links marked Could not verify?
Many sites refuse automated requests with HTTP 401, 403 or 429, even when the page works fine in a browser. LinkedIn, for example, often answers checkers with a non-standard 999. Calling those links broken would be a guess, so we label them separately and leave the verdict to you.
Does it crawl my whole website?
No. It checks every link on the single page you enter, which keeps the result fast and precise. To cover more of your site, run it on each of your main page templates.
Do broken links hurt SEO?
A few broken links will not sink a page, but they waste crawl time, lead the value of internal links into dead ends, and send visitors to error pages. Broken internal links are the ones to fix first, because they are fully in your control.
What is the difference between a 404 and a 410?
Both mean the page is not there. A 404 means not found, with no promise about the future. A 410 means gone on purpose. For a link on your own site the fix is the same: point it to the right page or remove it.
Link monitoring, included in Pro
Links break after you check them.
Pages you link to move, sites close and old addresses start forwarding somewhere else. Pro checks every link on every page of your site each week and emails you the new broken ones, with the page they are on.
Today
Every link works.
You fixed the broken ones from your last check.
Two weeks later
Someone else's page moves.
A source you cite changes its address. Your link now ends on a 404, on every page that uses it.
That week
Siteiz checks every link on every page.
Broken links, redirects that land on a home page, and links AI crawlers cannot see.
In your inbox
1 new broken link, and where.
The link, the page it sits on, and what to replace it with. Fixed in a minute, not found in a year.
No AI guesswork: the same rules engine as the free scan, run every week.
Subject
example.com: score 93, 1 new broken link
example.com
93 /100
New broken links
example.org/reports/2025-prices
on example.com/pricing and 3 more pages
Fixed and confirmed
2 broken links fixed
Open your report →
Working links are step one.
The free AI visibility scan checks whether AI crawlers can see, understand and cite the page those links point to, graded A to F with the exact fixes.