Overview
Introduction
Duplicate or near-duplicate content shows up in a lot of ordinary, non-nefarious situations - a page that was cloned as a starting template and never fully rewritten, an old draft that got half-updated, or a suspicion that a competitor or scraper lifted your copy. Before reaching for a heavyweight commercial plagiarism scanner, it's often faster to just paste both texts side by side and see exactly how much they overlap.
This tool computes an honest, explainable similarity score between two pasted texts using shingled word n-grams and the Jaccard index, and shows the specific overlapping phrases so you can judge for yourself whether the overlap is meaningful duplication or just shared boilerplate.
What Is Duplicate Content Checker?
A pairwise text-similarity comparator: paste any two blocks of text and get back a similarity percentage, a plain-language verdict, and the list of exact phrases both texts share.
It compares only the two texts you provide, directly against each other, entirely client-side. It is not a web-wide plagiarism or duplicate-content crawler - see the limitations below for why that's a meaningfully different (and much narrower) capability than a service like Copyscape.
How Duplicate Content Checker Works
Each text is lowercased and tokenized into words, then broken into overlapping shingles - by default, every run of 4 consecutive words - collected into a set of unique phrases per text. Text A's shingles are compared against Text B's shingle set to find exact matches.
The similarity percentage is the Jaccard index: the number of shared shingles divided by the total number of distinct shingles across both texts combined (shared + unique-to-A + unique-to-B). A verdict label (likely duplicate, partial overlap, largely different) is derived from fixed percentage thresholds, and the actual matched phrases are listed so you can inspect what's driving the score.
When To Use Duplicate Content Checker
When you suspect a competitor, content farm, or scraper copied a page's content, and want quick, concrete evidence of exactly which phrases match before pursuing anything further.
When checking an old page against a rewritten draft to confirm how much of the original text actually changed, versus how much was left untouched.
When auditing a set of near-identical template pages (location pages, product variants) to see how much copy is genuinely duplicated word-for-word versus meaningfully differentiated.
Often used alongside Keyword Density Checker and Meta Tag Checker.
Features
Advantages
- Uses a transparent, well-established algorithm (Jaccard similarity over word shingles) and shows the actual matched phrases, not just an opaque score - you can verify the result yourself instead of trusting a black box.
- Runs entirely in your browser on pasted text, so draft or embargoed copy never leaves your machine, unlike a hosted plagiarism-scanning service.
- Works on any two texts regardless of source - both don't need to be live web pages, so it's just as useful for comparing two local drafts.
Limitations
- This only compares the two texts you paste in, directly against each other - it does not search the web for duplicates elsewhere, unlike commercial tools like Copyscape. A page can score 0% here and still be a duplicate of something published elsewhere online.
- Exact-phrase shingle matching means a thoroughly paraphrased copy (same ideas, different wording throughout) will score low even though it may still be functionally duplicate content from an SEO standpoint.
- Shared boilerplate - navigation text, legal disclaimers, properly attributed quotations - inflates the score without indicating actual copying; always check the matched-phrases list rather than trusting the percentage alone.
Examples
Best Practices & Notes
Best Practices
- Read the matched-phrases list, not just the percentage - a 40% score made of ten scattered generic phrases means something very different from a 40% score that's one long, unbroken run of matched text.
- Paste comparable content on both sides (body text vs. body text, not one page's full HTML including its own navigation and footer against another's plain body text), since unrelated boilerplate on one side dilutes the score in either direction.
Developer Notes
Shingle size auto-reduces to the shorter of the two texts' word counts when either is shorter than the requested 4 words, so very short snippets (a headline, a meta description) still produce a result instead of erroring out with zero possible shingles.
Duplicate Content Checker Use Cases
- Investigating a suspected scraper or competitor copy by pasting your original alongside the suspected duplicate
- Confirming how much text actually changed between an old page and its rewritten replacement
- Auditing near-identical template pages (city/location pages, product variant pages) for how differentiated their copy really is
Common Mistakes
- Treating any nonzero similarity score as proof of plagiarism without reading which phrases actually matched - shared boilerplate or a quoted, attributed source can produce a moderate score honestly.
- Comparing a full HTML page (including nav, footer, and sidebar markup) against a competitor's plain body text, which skews the score with unrelated shared boilerplate rather than the content that actually matters.
Tips
- If two nearly-identical pages are legitimately your own (e.g. templated location or product pages), a high score here isn't inherently bad - it's just a prompt to make sure each page also has enough unique, differentiated content to be worth indexing separately.