How this tool works
This does not search the web. There is no index behind it, and no free tool has one — comparing your text against everything published would require a crawl of everything published. What it measures is repetition inside the text you paste: how often you reuse the same phrases and the same vocabulary.
Your text is split into overlapping three-word phrases and the distinct ones are counted, then weighted against the ratio of unique words to total words — roughly seventy per cent from the phrases, thirty from the vocabulary. A high score means you are not repeating yourself. It says nothing about whether anyone else has written the same thing.
Reading the score honestly
- It cannot detect copying: a passage lifted word for word from another site scores high here, because it is internally varied. Use this to catch your own repetition, not someone else's.
- Short text distorts it: under about two hundred words there are too few phrases for the ratio to mean anything. Paste whole articles rather than paragraphs.
- A low score points somewhere real: when it drops, the cause is usually one phrase you have leaned on — a product name, a stock transition — appearing in every second paragraph.
- Quoted material drags it down: block quotes, disclaimers and boilerplate count as repetition. Strip them out if you want a reading on your own prose.