Check a link preview without signing in

Why you ended up here

The platform-run debuggers each sit behind an account. Meta's Sharing Debugger needs a Facebook developer login. LinkedIn's Post Inspector needs a LinkedIn session. X's card validator is no longer the tool it was, and the advice pointing at it is mostly older than the change. None of that is unreasonable on their part — they are debugging their own crawler, not yours.

The gap is that most of the time you do not need their crawler. You need to know whether the tags on the page are right, which is a question about your HTML, and one you should be able to answer for any public URL — a client's, a competitor's, a staging domain — without logging in to a social network first.

What this cannot do: it cannot clear a platform's cache. Only the platform's own debugger can force a re-scrape, so once the audit confirms your markup is correct, that is still the step that makes an already-shared link update.

Five reasons a preview looks wrong

• The preview shows no image at all

  og:image is missing, or it is a relative path. Scrapers do not resolve relative image URLs — the tag has to carry a full https:// address.

• The preview shows an old image or an old title

  The platform cached the card the first time anyone shared the link, and it is still serving that copy. Your markup may already be correct. This is the single most common reason a fix appears not to have worked.

• The preview shows the site name instead of the page title

  og:title is absent, so the scraper fell back to the document title or the site name. The audit reports both separately, so you can see which one it actually found.

• The image is cropped strangely

  twitter:card is set to summary rather than summary_large_image, or the image's own aspect ratio does not match what the platform crops to.

• Nothing unfurls at all

  The page may be returning a non-HTML content type, redirecting somewhere the scraper will not follow, or blocking the scraper's user agent. The audit reports the status code and content type it received.

The order to check things in

1. 1. Confirm the tags are actually there. Run the audit above. If og:image or og:title is reported absent, nothing downstream matters until that is fixed.
2. 2. Confirm the image URL is absolute. A relative og:image is the defect that most often survives a visual check, because the image loads fine in your browser and not at all for a scraper.
3. 3. Only then blame the cache. If the markup is correct and the preview is still stale, the platform is serving an old copy and you need its debugger to force a refresh.