shared words appear in both texts, in order. Plain words are unique to that text.
Text A
Text B
This score reflects only the two texts above. It is not a plagiarism check against the web, and a low score does not prove two texts are unrelated, paraphrasing keeps the meaning while changing the words.
How it works
The Text Similarity Checker compares the two texts you paste, entirely in your browser. It tokenises both into words, finds the longest run of words they share in the same order (the longest common subsequence), and reports the overlap as a percentage.
- Similarity score = 2 × shared words ÷ (words in A + words in B), as a percentage. Identical texts score 100%.
- Highlights mark every shared word in both texts so you can see what matches and what is unique.
- Vocabulary overlap is a second view: the share of distinct words the two texts have in common, regardless of order.
Honest framing. This compares two texts against each other; it is not a web-wide plagiarism scan and cannot tell you if one text was copied from the internet. It also measures wording, not meaning, so a heavy paraphrase scores low even when the ideas match. Nothing is uploaded.
What you will see
- Similarity score
- The headline overlap percentage, with a verdict from Identical down to Very low overlap.
- Words in A / Words in B
- The word count of each text.
- Shared words
- How many words the two texts have in common, in sequence.
- Vocabulary overlap
- The share of distinct words common to both texts, ignoring order, a looser measure than the headline score.
- Highlighted panes
- Both texts side by side, with shared words marked so reused passages stand out.
Frequently asked questions
How does the text similarity checker work?
Paste a text into each box and click Compare texts. The tool finds the longest sequence of words the two share, in order, then expresses the overlap as a similarity percentage. It also highlights every shared word in both texts so you can see exactly what matches and what is unique.
Is this a plagiarism checker?
No, and we will be straight about that. This tool compares the two texts you paste against each other. It does not search Google, a journal database, or the wider web, so it cannot tell you whether a single text was copied from somewhere online. Use it to compare two drafts, a quote against its source, your text against a competitor's, or two documents you already have. A true web-wide plagiarism scan needs a paid search corpus we do not use here.
How is the similarity percentage calculated?
We tokenise both texts into words (lowercased, punctuation removed) and compute the longest common subsequence, the longest run of words that appears in the same order in both. The score is 2 × shared words ÷ (words in A + words in B), expressed as a percentage. Two identical texts score 100%; two texts with no words in common score 0%.
What do the highlights mean?
Highlighted words are the ones both texts share, in sequence. Un-highlighted words are unique to that text. Reading the two panes side by side shows at a glance which passages were reused and which were rewritten or added.
Will it detect paraphrasing or AI-rewritten text?
Only partially. Because the score is based on shared words, a heavy paraphrase that swaps most words will score low even though the meaning is the same, and an AI rewrite will usually score low too. The tool measures wording overlap, not meaning, so a low score does not prove two texts are unrelated.
Is my text uploaded or stored?
No. The entire comparison runs in your browser using JavaScript; neither text is sent to our servers or saved anywhere online. Close the tab and the text is gone.
Is there a length limit?
There is no hard limit, but very long texts (tens of thousands of words each) switch to a faster, order-insensitive comparison so your browser stays responsive. For typical drafts, articles, and essays the full word-order comparison runs instantly.