Paste two versions of any text and see every difference highlighted — green added, red removed, word- or line-level. Compared entirely in your browser, so contracts and drafts stay on your device.
groen = toegevoegd · rood = verwijderd. Vergeleken in je browser — contracten en concepten blijven op je apparaat.
What this free tool is great for: a quick, one-off job with no signup — it runs entirely in your browser, so nothing leaves your device and there's nothing to manage.
Its honest limit: it produces a one-off output — it won't store, track, brand or manage them at scale, and it can't tell you what happens after you share it.
Ask anyone to find the differences between two versions of a contract clause or an email draft and watch them read the same sentences four times, miss the changed number, and swear the texts are identical. Human reading is built for meaning, not comparison — we autocorrect small changes out of perception, which is precisely what makes manual diffing dangerous for exactly the documents where a changed word matters most. A diff tool compares mechanically: every added word green, every removed word red, nothing overlooked because nothing is read. This one runs entirely in your browser — worth emphasising, because the texts people most need to compare (contracts, offers, policies) are the ones that should never be pasted into a random website.
The two modes answer different questions. Word-level shows surgical edits inside sentences — ideal for prose, contracts and copy, where you want to see that "within 30 days" became "within 14 days" without the whole paragraph lighting up. Line-level treats each line as a unit — the right lens for code, config files, CSV rows and lists, where a line is the meaningful atom and word-level would produce noise. Rule of thumb: flowing text gets word-level, structured text gets line-level. For very large texts, line-level is also the practical choice — word-level comparison grows expensive quadratically, which is why the tool suggests switching when inputs get huge.
Green marks what the second version added; red-with-strikethrough marks what it lost. The skill is reading them as pairs: red directly beside green is usually a replacement (the interesting kind — a term changed), while lone green blocks are insertions and lone red blocks are deletions. Scan for the high-stakes categories first: numbers, dates, names, negations ("not", "no", "except") — the one-word changes that invert meaning. And read the stats line before the colours: "+2 added, −1 removed" on a three-page document tells you to hunt for needles; "+140, −92" tells you this was a rewrite and needs a full re-read, not a spot-check.
The highest-stakes diff in business life: a counterparty returns "the same contract with small changes". Trusting their summary of the changes is negotiating blind — the honest baseline is diffing their version against yours and reading every coloured mark. The same applies to updated terms of service, revised quotes, insurance policy renewals and any "v2" that arrives without a change list. Two habits complete the practice: always diff against the version you last approved (not the version they say they changed), and keep dated copies so that baseline exists. Five seconds of pasting beats an hour of side-by-side squinting — and beats a missed liability clause by considerably more.
Beyond adversarial reading, diffs answer everyday editorial questions. What did the editor actually change in my article? Paste both versions — every silent tweak becomes visible, which is both a quality check and the fastest way to learn from a good editor. Which copy did we actually publish versus what legal approved? Diff settles it in seconds. What changed between the email that converted and the one that flopped? The comparison turns vague intuition into a specific list of differences to test. Anywhere two versions of text exist and memory claims to know the difference, the diff knows better.
A diff shows that text changed, not whether the change matters — "shall" to "will" might be cosmetic or legally significant, and only a competent reader knows which. It compares character sequences, not meaning: a paragraph moved elsewhere shows as deleted-here-added-there rather than "moved", and reworded-but-equivalent sentences light up as fully changed. Formatting is invisible: bold, colours and layout live outside plain text, so a Word document's visual changes need Word's own review tools. Use the diff to guarantee you've seen every change; judging each one remains gloriously human work.
A diff is only as good as the versions you kept. The minimal discipline: date-stamped copies at every meaningful milestone (sent-to-client, received-back, approved), named so the sequence is obvious — the same sequential sanity we preach for invoices. For living documents, an "as approved" snapshot is the anchor everything else gets compared against. Teams that keep these habits resolve "who changed what" questions in seconds and without accusation — the diff is neutral in a way memories never are. Teams that don't keep them get to reconstruct history from email attachments, which is its own punishment.
A use that barely existed two years ago now dominates: comparing AI-generated versions. You asked a model to "tighten this paragraph" — what did it actually change? Paste both versions and every silent edit lights up, including the ones you didn't ask for: dropped nuances, altered numbers, softened claims. The same check protects prompt iterations (what did v2 of my prompt change in the output?) and editing workflows where an assistant rewrote customer-facing copy. AI editors are confident and occasionally wrong in quiet ways; the diff is the two-second audit that keeps the human genuinely in the loop instead of ceremonially — and it turns "the model changed something somewhere" into an itemised list you can accept or veto.
This tool compares text you can paste; much of what actually needs comparing arrives as PDFs — signed contracts, official quotes, scanned amendments — where copy-paste mangles layout and tables. That's where Foxit does more: native PDF comparison that highlights differences between two documents page by page, plus the editing, OCR and e-signing that contract workflows need around the comparison. Paste-able text gets diffed here, free and private; when the documents are PDFs with formatting that matters, compare them in the format they live in — the Acrobat-grade toolkit without the Acrobat price.
Green is text the second version added; red with strikethrough is text it removed. Red next to green is usually a replacement — the interesting kind. Check the +/− stats line first to gauge whether it was a tweak or a rewrite.
Word-level for prose and contracts (see surgical edits inside sentences); line-level for code, config, CSV and lists (the line is the meaningful unit) — and for very large texts, where word-level gets slow.
No — the comparison algorithm runs in your browser. That's the point: the texts people most need to diff (contracts, offers, policies) should never be pasted into a random server.
Paste the text content and yes — but formatting (bold, layout, tables) is invisible to a text diff. For PDFs where visual layout matters, use a native PDF-compare tool like Foxit's.
Blogger, teacher or toolmaker? Put this calculator on your own page — free forever, no strings. Copy the snippet below (the credit link is appreciated and keeps the tool free):
This tool is free and runs entirely in your browser. The link above is an affiliate link: we may earn a commission if you sign up, at no extra cost to you, and it never changes our honest take.