Compare two texts
Paste both versions and see exactly what changed, side by side, down to the individual words. The same algorithm git uses — not a line-by-line guess that falls apart when you add a line.
- Processed on your device
- No upload, no waiting
- No signup, no watermark
How it works
Paste both versions
Original on the left, new on the right. Or drop two files in.
Read the differences
Added, removed and changed lines are marked, and within a changed line the exact words are highlighted.
Ignore what does not matter
Capitalisation, spacing and blank lines can each be ignored — and the page tells you when that is the only reason two texts match.
Frequently asked questions
Are my documents uploaded?
No. Nothing is uploaded and no request is made — the counting happens inside this page, on your own device, and it keeps working with the network disconnected. That matters more than it looks for text: what people paste into an online counter is unpublished writing, legal drafts, student work and internal documents. For comparing two versions of a contract or a draft, that is the whole point.
Why is this better than comparing line by line?
Because a line-by-line walk breaks the moment anything moves. Add one line to the top of a file and every line below it is now on a different row, so a naive comparison reports the entire document as changed — which is useless. This uses Myers' algorithm, the one behind git diff and diff(1), which finds the smallest set of insertions and deletions that turns one text into the other. Add a line at the top, and it reports one added line.
Can it show which words changed, not just which lines?
Yes. When a line has been edited rather than added or removed, it is compared again word by word and only the words that actually changed are highlighted. On a long paragraph where someone changed three words, you see three words, not a wall of colour.
What do the "ignore" options do?
They change what counts as the same line. Ignoring capitalisation, or spacing, or blank lines is useful when comparing reformatted code or re-pasted text. One important detail: if two texts match ONLY because something is being ignored, the page says so explicitly rather than reporting them as identical — "identical" and "identical if you ignore case" are different facts, and confusing them is how someone ships a file they believed was unchanged.
Can I get the result as a patch?
Yes — the comparison can be copied as a unified diff, the same format git and email patches use, with a few lines of context around each change. That is the format to paste into a bug report.
Is there a size limit?
Your device's memory rather than an upload cap. Comparing two long documents is fast because the identical beginning and end are matched and skipped before the real work starts. Two texts that differ in thousands of separate places will be refused with an explanation rather than freezing the tab — a diff that large is not readable anyway.
Can it compare Word or PDF files?
Not directly. Convert them to text first — our PDF to Text and DOCX to TXT tools do that in the browser too — and paste the results here.