Compare two versions of a document to see what changed. The text of both PDFs is compared word by word, and additions and removals are highlighted on each page.
Drop the earlier file under Original and the later one under Revised, or click each box to browse. Either file can also come from Google Drive. For a password-protected PDF, you are asked for the password and the file is unlocked in your browser.
Click Compare. Pages are compared in order, page 1 with page 1 and so on. The bar at the top shows how many words were added and removed.
Each change is listed with its page number. Click one, or use Previous and Next, to jump to it in the text. Added words are green and underlined, removed words are red and struck through. Tick Show unchanged pages to see the full text.
More options has Ignore case, Ignore punctuation and Ignore line breaks. Use them when only the wording matters, for example when a file was re-typeset and the lines now break in different places.
The comparison uses the text inside the PDFs. A scan with no selectable text shows a note instead of results. Run it through OCR PDF first, then compare.
The comparison works on the text stored in each page, not on what's painted on screen. Usually they match. When they don't, the gap is useful.
Text covered by a white box is still text. If someone changed a figure by drawing a box over the old number and typing a new one on top, the comparison shows the new number as added and the old one as unchanged, because it's still in the file underneath. The same goes for text hidden behind an image, text in white on a white background, and the invisible layer that OCR adds to a scan.
This also lets you check a redaction. Compare the original with the redacted copy. If the words you meant to remove show as removed, they're gone from the file. If they show as unchanged, the redaction was only a black shape drawn on top. A page done with Redact PDF becomes an image, so all of its text shows as removed, not just the words under the box.
Each page's text is split wherever there's a space or a line end, and every piece is compared as one word. A few things fall out of that rule:
These add noise, so the word counts can run higher than the real edit.
Ignore punctuation removes dots, commas, currency signs, percent signs and similar symbols from every word before comparing. That includes the ones inside numbers. With it ticked, 4.2 and 42 count as the same, and so do $1,000.00 and $100,000, because both come out as 100000.
For prose that's fine. For anything with prices, rates, clause numbers or dates written as 01.02.2026, run the comparison once with the option off and read every change that involves a number. Ignore case is safer, but it treats a defined term like "Seller" the same as the ordinary word "seller", which can matter in a contract.
Headers and footers are text too. If the revised file is one page longer, a footer like "Page 3 of 12" becomes "Page 3 of 13" on every page, and each one appears in the list. Print dates, draft stamps and file paths in a footer do the same. Skip those entries and read the body text.
When the two files came from different programs, run both through PDF to Text first and skim the output. If the same paragraph comes out in a different order in each, the comparison will show it as moved text, and you'll know why before you start reading the list.
No fixed limit is set. Both files are processed in your browser, so very large files depend on your device's memory.
Yes. Long documents are easier to review on a larger screen.
Both files are compared on your device and are not uploaded. Their text stays in your browser.
