Paste the original version into the left editor and the modified version into the right, then press Compare Texts. Differences are highlighted in place: removed content in red on the original side, added content in green on the modified side. Nothing is uploaded and there is no sign-up.
Split view puts the two versions side by side, which is the clearest way to see what a passage looked like before and after. Unified view interleaves them into a single column in the style of a Git diff, which reads better when changes are scattered through a long document. Switch between them at any time — the comparison does not need re-running.
Word-level is the sensible default for prose: a changed word is marked as one change rather than a scatter of altered letters. Character-level goes finer and is worth switching to when the edits are small and precise — a digit in an ID, a flipped letter in a typo, a changed file extension — where word-level would just mark the whole token.
The statistics panel counts changes two ways. The first section tracks words or characters — how many were removed, how many added, the net change, and the total affected. The second tracks lines removed, added, modified, and changed in total. The net figure is the useful one for judging whether an edit grew or shrank the text; the total is the better measure of how much was touched.
In a long document the differences are rarely near each other. The up and down arrows step through them one at a time, with a counter showing which change you are on, and the keyboard shortcuts do the same — arrow up and down, or j and k if you are used to Vim. This is usually faster than scrolling and hunting for highlights.
Text mode compares exactly as written, which is what most code and configuration wants: a changed value, a moved line or an altered indent shows up as written rather than being normalised away. Structured formats are better served by their own tools, linked below — those compare the data rather than the characters, so reformatting and reordered keys stop counting as changes.
The comparison runs in JavaScript on your own machine. There is no backend to send anything to — the whole site is a static export — so contracts, credentials, patient notes, and unreleased code can be compared without any of it crossing the network. That is a genuine difference from diff tools that upload both sides to a server before showing you the result.
Paste the original into the left box and the modified version into the right, then press Compare Texts. Removed content is highlighted in red and added content in green, with line numbers so you can find each change in the source.
Split view shows the two versions side by side, which makes before-and-after easy to read. Unified view merges them into one column in the style of a Git diff, which is easier to scan when changes are spread through a long document.
Word-level marks a changed word as a single difference and suits ordinary prose. Character-level compares letter by letter, which is more precise for small edits like a changed digit or a typo, but noisier on rewritten sentences.
No. The comparison runs entirely in your browser. Easy Formatter is a static site with no backend, so neither version of your text is transmitted anywhere and nothing is stored.
Yes. Text mode reports exactly what changed in the characters, which suits most code and configuration. For structured formats there are dedicated tools linked below, which compare the data rather than the characters so that reformatting does not register as a change.
Yes. Copy the differences to your clipboard, or download them as a text file to attach to a review, a ticket, or a record of what changed.
Use the up and down arrow buttons, or the keyboard: arrow up and arrow down, or j and k. A counter shows which change you are currently on, so you can work through a long document without scrolling.
There is no imposed limit. Because the work happens in your browser rather than on a server, the practical ceiling is your own device memory rather than an upload cap or a rate limit.