What happens when you click Check
Unlike the generator tools on this site, this one does not run in your browser alone. Your URL is sent to this site's own /api/page-meta endpoint, a Cloudflare Pages Function that fetches the page server-side, follows redirects and parses the returned HTML with HTMLRewriter. The results come back as structured data and are rendered in sections. Because the server does the fetching, you avoid browser cross-origin limits and can inspect any public page. The same request is also guarded so it cannot be pointed at private or internal addresses.
What you get back
- Google SERP preview using the page's title, description and hostname.
- Title and description each with a character count, a pixel width measured in your browser, and a "likely truncated" flag past 600px (title) or 920px (description). Colors are red when over the pixel limit, amber when outside the suggested range of 30 to 60 characters for titles or 70 to 155 for descriptions, and green otherwise.
- Core tags: canonical, robots, lang, charset, viewport, favicon, the final URL after redirects and the HTTP status code.
- H1 headings: the first three found on the page.
- Open Graph and Twitter Card tables and an Other meta tags table for everything else.
- Social card preview that falls back from
og:titleandog:descriptionto the plain title and description when the OG versions are missing. The image URL is displayed as text, not loaded.
Worked example
Checking a blog post might return a title of 78 characters, flagged red as likely truncated, a canonical pointing at the same URL without a trailing query string, a robots value of index, follow, a status of 200, and two H1s. That tells you three things to fix: shorten the title, remove the duplicate H1, and add an og:image because the OG table has none. A final URL different from the one you typed means the page redirected, for example from http to https or to a trailing-slash version, which is worth confirming is intended.
Reading the key tags
| Field | What to look for |
|---|---|
| Canonical | Absolute URL, normally pointing at itself |
| Robots | Unexpected noindex on a page you want ranked |
| HTTP status | 200 expected; 3xx, 4xx or 5xx need investigating |
| Viewport | Present for mobile-friendly rendering |
| H1 | Exactly one, matching the page topic |
| og:image | Present, ideally 1200x630 |
Limits and common mistakes
- JavaScript-rendered tags are invisible here. The tool reads the initial server HTML. If a framework injects the title or canonical after loading, they appear missing even though Googlebot, which renders JavaScript, may see them. Compare with a "view source" check.
- Some sites block automated requests. A firewall or bot protection may return an error or a challenge page instead of your content, so a failure does not always mean your tags are wrong.
- Only the first three H1s are listed. Several H1s is a sign to review your template, though it is not a penalty on its own.
- Pixel widths are approximations measured with Arial, as in the SERP snippet checker.
For a step-by-step audit routine, see the guide to checking meta tags. To dig deeper into duplicates, use the canonical checker, and to fix what you find, build replacements with the meta tag generator.