Content

Duplicate Content Checker

See exactly how much two pages share

In build

Paste two URLs or two blocks of text and get a similarity score for the main content, the passages that match word for word, the share of each page the match accounts for, and the canonical each URL declares.

Get early access

What it returns

A similarity percentage for the main content of each page, with navigation, sidebars, footers and template boilerplate removed before anything is compared.
The matching passages highlighted in both texts, so you can see whether the overlap is a shared intro or the entire body.
The share of each page the match accounts for, which is the number that decides anything: one repeated paragraph is trivial on a long page and almost the whole thing on a short one.
The canonical each URL declares, whether the two agree, and a note when one of them already redirects to the other.

When this matters

Two category pages were built from the same template two years apart, and now one ranks for nothing while the other drifts around. Nobody has read them side by side because they are long and most of the length is an identical specification block. There is no duplicate content penalty to worry about here. What actually happens is that Google picks one URL to represent the set and filters the rest out of results, so the question is not whether you will be punished but which URL is being dropped and whether it is the one you would have chosen.

How to answer this today

Site audit finds this at scale: it crawls every page and clusters the ones whose content is near-identical, which is how a templated section with four hundred thin variants gets discovered by someone who was not looking for it. Comparing exactly two pages you already suspect, one of which may not be on your site at all, is faster and is the case a crawl is bad at.

Site audit

Why this is harder than it looks

There is no duplicate content penalty

The real mechanism is canonical selection. Google groups URLs it considers equivalent, picks one to show, and the others stop appearing. That looks like a penalty and is not one, and the difference matters because the remedy is to consolidate or differentiate rather than to fear a punishment that was never applied.

Boilerplate dominates the score

Header, nav, sidebar, footer and legal text can be most of the markup on a page. Compare raw HTML and two completely unrelated pages on the same template score as near-duplicates. Main content has to be isolated first, and that extraction is the part that decides whether the number means anything.

Near-duplicate is the case that matters

The damaging pattern is rarely a byte-for-byte copy. It is the same page with a city name swapped, sentences reordered, or one product number changed. Exact matching misses all of it, so comparison has to run on normalised text and report partial overlap rather than a yes or no.

You suggest the canonical, Google chooses it

A canonical tag is one signal among redirects, internal links, sitemaps and page similarity. Google can and does select a different URL. A comparison tool can show what you declared and whether it is consistent. The URL Google actually chose is only visible by inspecting the URL in Search Console.