Open Graph Preview Checker
Inspect declared social metadata, preserve Open Graph arrays and fallbacks, and build only an approximate preview.
Add the real page URL to resolve relative social/canonical URLs. No request is made.
Only the supplied HTML is parsed. No page, image, redirect, HTTP header, or social platform's cached preview is fetched.
Open Graph properties, image groups, X card fields, fallbacks, and conflicts will appear here.
A Fallback Can Make the Preview Look Fine While the Open Graph Markup Is Still Incomplete
A normal HTML title and meta description can supply text for this approximate card. Some real services also use fallbacks. But seeing a readable title in a preview should not be confused with actually declaring the four Open Graph core properties.
The report therefore evaluates og:title, og:type, og:image, and og:url separately even when the card has something reasonable to display.
Repeated og:image Is an Array, Not Automatically a Duplicate Error
Open Graph explicitly allows multiple values by repeating the same root tag, and gives preference to the first tag when values conflict. That makes two og:image tags different from accidentally emitting two contradictory og:titlevalues.
Repeated image roots remain distinct candidates, with width, height, type, and alt properties attached to the image that precedes them in source order.
Structured Properties Can Quietly Attach to the Wrong Image
<meta property="og:image" content="first.jpg"> <meta property="og:image:width" content="1200"> <meta property="og:image" content="second.jpg"> <meta property="og:image:alt" content="Second image">
In this example, the width belongs to the first image and the alt text belongs to the second. Moving every structured property to the bottom changes that relationship even if every individual tag still looks valid.
Pasted HTML Cannot Prove Whether the Image URL Actually Works
A syntactically valid og:image can still return 404, redirect through an inaccessible host, block a social crawler, serve an unexpected content type, or be cached under an older version. None of those facts exists in pasted HTML.
Inspect the declarations here, then use the target platform's live debugging or card refresh workflow to verify what its crawler can actually fetch.
canonical and og:url Can Differ, but the Difference Should Be Deliberate
Search canonicalization and Open Graph identity are separate mechanisms. A canonical link tells search systems which duplicate URL you prefer as representative; og:url identifies the graph object for social metadata.
On an ordinary article or product page they often point to the same stable URL. A mismatch is therefore worth reviewing, not because the standards require equality, but because accidental campaign, hostname, protocol, or slash differences are common.
X-Specific Copy Is Not Necessarily a Conflict
A page can deliberately use a shorter twitter:title or different social description while preserving the same underlying page identity. The report shows the difference as a note rather than forcing Open Graph and X values to be identical.