How to compare two texts
- Paste the first version into the Original box.
- Paste the second version into the Changed box.
- Optionally tick Ignore case or Ignore leading/trailing spaces so cosmetic differences do not show up.
- Click Compare. The result fills in below: green
+lines were added, red-lines were removed, and plain lines are unchanged. On a line that was edited rather than fully replaced, only the specific words that changed are highlighted, the rest of the line stays plain. - Read the counts at the top, including the similarity percentage, then press Copy diff to grab a plain-text version with
+,-and space prefixes.
How the line and word diff works
Under the hood the tool runs a classic longest-common-subsequence (LCS) diff over the lines of your two texts, the same idea behind the diffs you see in Git and code review tools. It finds the longest set of lines that appear, in order, in both versions, and treats everything else as an edit. Lines that exist only in the Original are reported as removed, lines that exist only in the Changed text are reported as added, and the shared backbone is shown as unchanged. Because it works on whole lines, moving a paragraph shows up as one block removed and one block added, which is usually exactly how you want to read a revision.
When a removed line and an added line sit next to each other, the same edited line on both sides, the tool runs a second LCS pass over the words in that line, so only the exact words that changed get wrapped in a highlight and everything else renders as plain text. The similarity percentage above the result is built from that same word comparison: it is the share of words the two texts have in common, in order. It is a structural comparison, not a meaning comparison, so two passages that say the same thing in very different words will still score low.
When a text diff is handy
A line-based diff is perfect for spotting edits between two drafts of an email, contract or article, checking what a teammate changed in a config file or list, confirming a copy-paste came through intact, or reviewing the before-and-after of generated output. The addition and deletion counts give you a quick sense of how big a change is at a glance, a one-line tweak versus a wholesale rewrite, before you read a single line. Your original text stays in the input boxes, so you can toggle the case and whitespace options and compare again without losing anything.
Frequently asked questions
Does it diff word by word or line by line?
Both. It first compares whole lines to find what was added, removed or changed. Then, for any line that exists on both sides but was edited, it runs a second word-level pass so only the words that actually changed are highlighted, not the whole line.
What do the counts mean?
Added is the number of lines present only in the Changed text, Removed is the number present only in the Original, and Unchanged is the count of lines that matched in both.
What does the similarity percentage mean?
It measures how many words the two texts have in common, in order, not whether they mean the same thing. It is a structural comparison, not a plagiarism or meaning check, so two passages that say the same thing in different words will still score low.
Related: find & replace text · sort text lines · clean up messy text