Load an original and a revised PDF and see precisely which lines were added or removed, with page numbers, a change summary and a report you can send on.
Comparison works on text in reading order, so a line present in both documents anchors the alignment and everything else falls out as an insertion or a deletion. A sentence that moved therefore appears twice: removed in one place, added in another. That is how diffs work, and it is usually what you want to see.
Change a margin, a font size or a page break, and every paragraph re-wraps. The words are identical but the lines are not, and a strict comparison would flag the entire document. Normalising spacing before comparing strips that noise out and leaves the substantive edits.
It compares words, not appearance. A logo swap, a colour change, a moved table or a different signature block are invisible to it. For those, export both versions with PDF to PNG and compare the images page by page.
A scanned PDF contains no text, so there is nothing to compare. Run each one through OCR first and compare the searchable exports — bearing in mind that OCR differences can produce diffs of their own.
The text, line by line, in reading order — the same longest-common-subsequence approach git uses. Layout, fonts, images and colours are not compared, so a re-styled document with identical wording shows as unchanged.
No. Both files are read and compared inside your browser, which matters when the two versions are contract drafts.
If a paragraph was re-wrapped, its lines break at different points and no longer match. Leave “Ignore whitespace and re-wrapping” on — it normalises spacing before comparing, which removes most of that noise.
Only if they have a text layer. Run each scan through the OCR tool first, export the searchable version, then compare those.
A plain-text summary — both filenames, page counts, totals, the comparison options used — followed by every added and removed line tagged with its page number. It is designed to be pasted into an email or a ticket.