Text Similarity Checker
Paste two texts to see how much of their wording overlaps, measured by shared vocabulary — useful as a quick sanity check, not a plagiarism verdict.
What this percentage actually means
This tool measures Jaccard similarity: the number of unique words the two texts share, divided by the total number of unique words across both. It compares vocabulary, not meaning or sentence order — two texts can score low here while saying the same thing in different words, or score higher purely by sharing common topic terms.
What this is — and isn't
- Useful for a quick check of how closely two drafts, quotes, or paraphrases overlap in wording
- Not a plagiarism detector — it has no database of published text to compare against, only the two texts you provide
- Case-insensitive and ignores punctuation, so rewording without changing vocabulary can still score high
Frequently asked questions
Is this a plagiarism checker?
No — it only compares the two texts you paste in against each other, with no database of published books, articles, or websites to check against. For genuine plagiarism detection you'd need a service with an indexed content database.
What counts as "similar" — the same meaning or the same words?
This tool measures shared vocabulary (Jaccard similarity), not meaning. Two texts saying the same thing in completely different words can score low, while two texts sharing common topic terminology can score higher even if their overall meaning differs.
Why does the similarity score seem low even though the texts are clearly related?
If the texts are paraphrased or reworded significantly, they may share few exact words even while conveying the same idea — since this tool measures word overlap, not semantic meaning, heavily reworded text will naturally score lower.
Can I use this to check if I copied someone else's writing by accident?
It can help you spot close wording overlap between your draft and a source you paste in, but it's not a substitute for a proper plagiarism-detection service that checks against a wide index of published content.