Skip to main content
ToolsHub

Text Similarity Checker

Paste two passages to see how similar they are, with cosine and Jaccard scores plus shared and unique words. Useful for spotting duplicate content — all computed in your browser.

Updated

Files never leave your browser
Cosine similarity
55%
Jaccard similarity
60%
Shared words
6

Shared: quick, brown, fox, over, lazy, dog

How to use Text Similarity Checker

Compare two texts side-by-side and measure how similar they are using cosine and Jaccard similarity algorithms — all computed in your browser with no text ever uploaded to a server. Useful for detecting duplicate content, checking paraphrase overlap, comparing contract revisions, and identifying near-duplicate submissions. Results include a percentage similarity score for each algorithm, a list of shared terms, and vocabulary unique to each passage. Ideal for writers, editors, researchers, and SEO professionals.

  1. Paste the first text into the left input field.
  2. Paste the second text into the right input field.
  3. Click "Compare" to run both cosine and Jaccard similarity calculations.
  4. Review the similarity percentage and the breakdown of shared versus unique words.
  5. Use the highlighted diff view to spot paraphrased sentences or substituted phrasing.

Your data never leaves your device — 100% private processing.

Cosine similarity explained

Cosine similarity measures the angle between two text vectors in a high-dimensional word-frequency space. Each document is represented as a vector of term frequencies, and the cosine of the angle between them — (A·B) divided by (||A|| × ||B||) — gives a value between 0 (no word overlap) and 1 (identical word distribution). Because cosine similarity is length-agnostic, it works well even when one text is much longer than the other, making it the preferred metric for content duplication detection and document comparison.

Cosine vs. Jaccard comparison
MetricFormulaLength sensitivityBest for
Cosine similarity(A·B) / (||A|| × ||B||)Low (length-agnostic)Documents of unequal length
Jaccard similarity|A ∩ B| / |A ∪ B|HighStrict vocabulary overlap checks

Jaccard similarity and when to use it

Jaccard similarity divides the number of unique words shared by both texts by the total number of unique words across both texts: |A ∩ B| / |A ∪ B|. A score of 1.0 means both texts use exactly the same vocabulary; 0 means no words in common. Jaccard is more sensitive to vocabulary differences than cosine similarity but less sensitive to term frequency. Use Jaccard for strict vocabulary-overlap checks and cosine when comparing documents of significantly different lengths.

Glossary

Cosine similarity
A metric measuring the cosine of the angle between two document term-frequency vectors, ranging from 0 (unrelated) to 1 (identical distribution), regardless of document length.
Jaccard similarity
The ratio of the intersection size to the union size of two word sets, measuring strict vocabulary overlap between 0 (no shared words) and 1 (identical vocabulary).
Term frequency (TF)
The count of how many times a word appears in a document, used as the basis for building vector representations of text for similarity comparison.
Tokenization
The process of splitting a text into individual words or tokens, typically lowercased and stripped of punctuation, before similarity algorithms are applied.
Near-duplicate
A document that shares most of its content with another — usually the result of paraphrasing, synonym substitution, or minor edits — detected by high similarity scores.

Related reading

Free · No spam

Get weekly tool tips & updates

New tools, power-user tips, and productivity hacks — delivered free every Friday.

No spam, ever. Unsubscribe with one click.

Related AI & Writing

Explore all AI & Writing.