What Websites Broken Link Checker does
Websites Broken Link Checker fetches one source page, extracts normalized internal links, and selects the first thirty for status checks. It sends HEAD where supported and falls back to GET for servers that reject HEAD, processing small batches to keep the crawl bounded.
A destination is marked broken when its final response is below 200, at least 400, or the request fails. Redirect responses that finish below 400 are not marked broken. This definition is useful for a technical snapshot but cannot determine whether destination content is correct, a redirect is intentional, or a transient failure will persist.
How to use Websites Broken Link Checker
- Enter the full public URL of the source page.
- Select Check page links to fetch the page and discover same-host destinations.
- Review discovered and checked counts, the truncation flag, final statuses, URLs, and error codes.
- Open or retest flagged links before editing content, redirects, or server configuration.
Important details
The crawl starts and ends at one source page
PagesTools discovers links in the submitted page but does not recursively crawl linked pages for more links. External-host destinations are not checked by this tool.
No more than thirty links are checked
The first thirty normalized same-host links are selected. discoveredInternalLinks and truncated show when the page contained more candidates than the report tested.
HEAD can fall back to GET
A lightweight HEAD request is attempted first. When a destination responds with 405 or 501, PagesTools retries with GET so servers without HEAD support are not automatically labeled broken.
Practical uses
- Find obvious internal 404 or 500 responses on a recently edited page.
- Review link outcomes after a URL migration or redirect update.
- Collect a small same-host status report before a manual content audit.
Privacy and limitations
The source URL is sent to PagesTools, whose server fetches it and up to thirty same-host destinations. Those servers can log each request. Avoid private or signed URLs; the JSON report uses no-store cache headers.
The tool checks one source page, at most thirty same-host links, and no recursive or external links. JavaScript-added links are absent, authenticated destinations are unsupported, and rate limits, bot defenses, or transient responses can create misleading failures.