Content Similarity Checker
How much text two pages share
What Is the Content Similarity Checker?
The Content Similarity Checker reads the main text of two pages and compares them five words at a time. It reports how similar they are overall, how much of the shorter page appears in the longer one, the longest passages they share and where each sits, how much vocabulary they have in common and the words only one of them uses, then gives a plain verdict from distinct to near-duplicate.
Key Features
Five Words at a Time
Each page's text is lowercased, stripped of punctuation and accents, and cut into overlapping five-word runs. Similarity is the share of all distinct runs that appear on both pages. Containment divides the shared runs by the shorter page's runs, which catches what similarity misses: a short page copied whole into a long one.
Reading the Verdict
Under 5% similar is distinct content. Up to 20% is usually boilerplate, quotes or standard wording. Past 20%, or once 40% of the shorter page appears in the longer, long stretches are shared and worth a look. At 50% search engines will likely show only one of the pages, and when most of the shorter page sits inside the longer one too, it's effectively a copy.
Navigation Stays Out
Both pages are read for their main content, so menus, headers and footers are stripped before anything is compared. A little boilerplate can survive, like a standard disclaimer, and it shows up as short passages you'll recognise and can ignore.
Common Use Cases
- Checking whether two of your pages compete as duplicates
- Seeing how much of your text a scraped or syndicated copy reuses
- Making sure template-built location or product pages say enough that differs
- Comparing a rewrite with the original before publishing
Frequently Asked Questions
Is the Content Similarity Checker free?
Yes. It's free with a ProxyHorizon account, and every account gets 25 runs a day shared across all our data tools. You can read about the tool without signing in, but running it needs an account because each run fetches live data.
Is this a plagiarism checker?
Not in the usual sense. It compares two pages you choose rather than searching the web for copies. When you know where a copy might be, it shows exactly which passages were reused and how much of each page they make up.
Why do two different pages still share passages?
Shared wording is normal. Disclaimers, quotes, product names and set phrases all repeat. Short shared passages are rarely a problem; a high similarity score, or one long passage that makes up a big share of a page, is what deserves attention.
Does it work in other languages?
Yes, for languages that put spaces between words. Words are split on letters and numbers in any script and accents are ignored, so Spanish or German pages compare as well as English ones. Chinese, Japanese and other languages written without spaces aren't split into words, so their scores aren't reliable.
Do you keep the pages?
For 24 hours. Each page read is stored for a day and shared by the page comparison, landing page comparison and similarity tools, so comparing the same page again, against anything, is instant. It only holds public pages. Choose "Read both pages again now" for fresh reads. We also log which tool you ran and when, to apply the daily limit.
Monitoring competitors at scale?
Proxies for market research