Social Preview Checker

How do you preview a link before the page is live?

You type the tags in rather than pointing a tool at a URL. That is the whole difference between this checker and every official debugger: Facebook’s Sharing Debugger and LinkedIn’s Post Inspector both need a URL that is already public, which means the first time you see your card is after the page is shipped and someone has already shared it badly. Here you paste the title, description and image URL you are about to write into the page, and read the three renderings side by side.

Once the page is live, use the official debuggers as well — they see what the crawler actually got, including caching and fetch failures this tool cannot know about. The two jobs are different: preview first, verify after.

What are the correct sizes and limits for a share image?

1200 × 630 pixels, under 8 MB, at an absolute HTTPS URL. Those are the numbers that matter, and they come from Facebook’s own sharing documentation rather than from received wisdom. Here is the full set, with the two places where the platform publishes no number at all called out honestly.

All values read from the platforms’ own documentation on 25 August 2026. X’s card image specifications are not included because its card documentation now sits behind the developer platform and we will not quote numbers we cannot open.

How long should your title and description actually be?

There is no published limit for either — and that is the single most useful fact on this page, because almost every article about metadata states one as if it were a rule. Google’s documentation says the title link “is truncated in Google Search results as needed, typically to fit the device width”, and that there is “no limit on how long a meta description can be”, with the same width-based truncation applied to the snippet.

So the familiar 60 characters for a title and 160 for a description are conventions, not limits: they are roughly what survives on a desktop result. They remain the right defaults, which is why our meta tag generator counts against them. But since truncation follows device width, the practical rule is to front-load the meaning: put the words that decide the click in the first 40 characters, and leave your brand name at the end where losing it costs nothing.

Will Google even show the title and description you wrote?

Often, but not always, and Google is explicit about it. Title link generation is “completely automated and takes into account both the content of a page and references to it that appear on the web”, and Google rewrites your title when it judges it inaccurate, obsolete, written in a different language from the page, repeated boilerplate across many pages, or missing information that is prominent on the page. Snippets are “primarily created from the page content itself”, and your meta description is used only when Google thinks it describes the page better than the page does.

The practical consequence is the opposite of what most SEO advice implies. You are not writing a title to be displayed verbatim; you are writing one accurate and specific enough that Google has no reason to replace it. Boilerplate like “Home | Acme” on forty pages is the most reliable way to get rewritten.

Why is my link preview blank or showing the wrong image?

A relative og:image path causes this more often than every other reason combined. Crawlers do not resolve relative URLs — /og.png is invisible to them, and only https://yourstartup.com/og.png works. The next most common causes, in the order we actually hit them:

• The image sits behind authentication or on a staging host that is password-protected.
• The image or the page is disallowed to crawlers in robots.txt.
• The file is over the 8 MB maximum, or served with an encoding the crawler does not accept — Facebook’s crawler takes gzip and deflate only.
• The tags are injected by JavaScript after load, so the crawler sees a head with nothing in it.
• The platform cached an earlier scrape. Re-fetch it in Facebook’s Sharing Debugger or LinkedIn’s Post Inspector, or change the image filename to sidestep the cache entirely.

Which preview tool should you use, and when?

Use this one while writing, and the official debuggers after deploying. They answer different questions. Facebook’s Sharing Debugger shows what the crawler actually fetched and lets you force a re-scrape, which is the only fix for a cached old image. LinkedIn’s Post Inspector does the same for LinkedIn. X no longer offers a public card validator — the old cards-dev.twitter.com URL redirects to an x.com login — so for X the preview-first workflow is not a convenience, it is the only option left.

The five-minute metadata checklist

Do this once per important page and you are ahead of most sites: write a title that is specific to the page and front-loaded; write a description that reads like a reason to click rather than a summary; set an absolute, HTTPS og:image at 1200 × 630; set twitter:card to summary_large_image; add a self-referencing canonical URL; and read the whole thing back at phone width before you ship it.

When the tags are written, generate the full block with our meta tag generator, give language models a readable map of the site with the llms.txt generator, and see the rest of the free founder tools. Perfect metadata on a page nobody links to still ranks nowhere, so once the page is live, work through how to get backlinks for a new website and list your startup on Launchory for a crawlable profile and a dofollow link.

Frequently asked questions