Text Diff Checker
Paste two versions of a text and see exactly what changed. Differences are highlighted by line, and within a changed line you can drill down to the word or character level. Nothing is uploaded.
Fill in both sides to compare.
Comparison runs entirely in your browser — safe for contracts and configs.
How to use
- Paste the original text on the left and the changed text on the right.
- Pick a granularity — lines for prose and code, words or characters for fine edits.
- Switch between the side-by-side view and the unified view, whichever you read more easily.
- Turn on “Ignore case” or the whitespace options if those differences should not count, then copy the patch.
Details
When something has changed, the useful question is rarely “did anything change” but “what exactly changed, and is it what I intended”. A good diff answers that by keeping unchanged content out of the way and showing only the lines that moved, and by pairing a deletion with the addition that replaced it so a rewritten line reads as one change rather than two unrelated ones.
Comparison runs entirely in your browser, which matters for the kinds of text people usually diff: contracts, configuration files, environment variables and API responses — material that often contains credentials or personal data and should not be pasted into a server-side service.
FAQ
Does “ignore case” rewrite the text that is displayed?
No. Comparison is performed on normalised copies, but everything shown on screen — and everything exported in the patch — is your original text, with its original capitalisation and spacing. The options only change which lines are judged to be equal.
What is the difference between the two views?
The side-by-side view puts the original and the changed text in two aligned columns, which makes it easy to see what a line became. The unified view interleaves them in a single column with + and − markers, which reads like the output of git diff and is more compact on a narrow screen.
Can I use the exported patch with git?
Yes. The unified view can be copied or downloaded as a .diff file. A leading separator line that some diff tools print is removed on purpose, because git and patch reject a file that starts with it. The remaining structure — file-agnostic hunks with @@ headers, context lines and +/− markers — is standard.
Why does it fall back to line comparison on some inputs?
Word- and character-level comparison is quadratic in the length of the lines involved. On documents with extremely long single lines — a minified JavaScript bundle or a single-line JSON file — that would become slow and memory-hungry, so the tool detects the situation and switches to line-by-line comparison, telling you that it did.