Duplicate Content Checker

How this works

We compute two signals and report the higher one. Phrase match breaks both texts into overlapping 5-word phrases and measures how many they share — precise for verbatim or near-verbatim copies, the same shingling approach crawlers use. Word overlapcompares individual word usage — it catches "spun" content that swaps in synonyms and reorders clauses, which can otherwise slip past phrase matching with a deceptively low score. It's a heuristic, not Google's verdict — always confirm impact in Search Console's Pages report.

How we tested this

• Identical and unrelated text — confirmed identical input scores 100% and two completely unrelated passages score 0%, at both signal levels.
• Spun content— we found that phrase-matching alone badly under-scored classic "spun" duplicates (a passage rewritten with synonyms scored just 6% on phrase match alone, in the "likely unique" band) even though it's the same content in the same structure. That's the reason the word-overlap signal exists — the same pair scores 63% on word overlap, correctly surfacing it as worth a canonical.
• False-positive check— tested two independently-written passages covering the same SEO concept in different words and structure (the kind of content that shouldn't be flagged): word overlap stayed at 15%, comfortably in the "likely unique" band, confirming shared topic vocabulary alone doesn't trigger a false duplicate verdict.
• Edge cases — empty inputs, text shorter than the 5-word shingle window, and case/punctuation-only differences (correctly scored 100% after normalization) all handled without errors.