How to use
- Open the page in your browser and press
Ctrl+U(Option+Command+Uon a Mac) to view its source. Copy it all, paste it into the box above and the results appear straight away. A partial paste works too, even just the<head>. - If you want to know whether the canonical tag matches the page address, type the address in Page URL.
- Read the error, warning and note counts and the finding for each item. Errors can directly affect search visibility, warnings are worth fixing, and notes depend on your situation.
- Use Copy report to take the results as plain text into a ticket or a note.
What it reads
| Item | What is checked |
|---|---|
title |
Present, empty, more than one, character count and display width |
meta description |
Same as above |
meta robots, googlebot and others |
noindex, nofollow, none directives |
link rel="canonical" |
Count, relative path, match with the page URL and og:url |
viewport, charset, html lang |
Whether they are declared |
link rel="alternate" hreflang |
Per-language URLs, presence of x-default |
og: and twitter: tags |
Title, description, image, URL, and whether the image is a full URL |
Icon link tags |
icon, apple-touch-icon |
| JSON-LD | List of @type values (including inside @graph), JSON syntax errors |
h1 count |
None, or several |
Characters versus display width
Search engines cut titles and descriptions by pixel width, not by character count, and they publish no official character limit. So the tool shows the character count together with a display width, a simple measure that counts Korean, Chinese and Japanese characters as twice the width of Latin ones. The guide ranges are the ones commonly used in the industry: width 30-60 for titles and 70-160 for descriptions. For example, Free Online Meta Tag Checker for Pages is 38 characters, width 38, so it is marked good, while a 31-character Korean title has width 62 and is marked long. Going outside the range does not by itself lower rankings; it is a way to anticipate truncation.
Reading the results
- The tool only reads the tags in the pasted source. Tags that scripts add after the page loads may be missing from view-source. For such pages, copy the
<html>element from the Elements tab of the developer tools instead. noindexis shown as an error, but login and checkout pages often carry it on purpose. Just confirm it is intended.- A page blocked in
robots.txtcannot have itsnoindextag read by search engines. Review both settings together with the robots.txt Generator. - To write missing tags use the Meta Tag Generator, to see share cards use the OG Preview, for heading structure the Heading Checker, and for the list of links the Link Extractor.
Frequently asked questions
Can I just enter a URL?
No. The tool does not fetch pages by address. It reads only the HTML source you paste, which you can copy from view-source (Ctrl+U).
How long should a title and description be?
There is no official limit. The tool uses display width 30-60 for titles and 70-160 for descriptions (a Korean character counts as width 2) as a guide. Putting key words first keeps the meaning if the text is cut.
The canonical exists but it says it differs.
It tells you when the canonical address is not the same as the page URL you entered. That is normal if you deliberately point parameterised URLs to a main address. A trailing slash, anything after # and host capitalisation are treated as the same address.
Why am I warned that og:image is a relative path?
Sharing services fetch the image from outside your page, so they cannot tell which site /og.png belongs to. Use a full address that starts with https://.
Where does the pasted source go?
The text you enter is processed only in your browser and never sent to a server.