Compare PDF

See the differences between two PDF files side by side.

or drop PDF here

Runs in your browser — your file never leaves your device.

See exactly what changed between two versions of a PDF. Compare PDF walks both documents page by page and compares each pair twice: pixel by pixel, which draws a red outline around every region that moved or changed, and word by word, which lists the text that used to be there and the text that replaced it. The result is a single self-contained HTML report you can open in any browser, offline, print, or send on. Both files are read on your device.

How to use Compare PDF

Click "Select PDF files" and add exactly two PDFs — the original first, the revised copy second.
Choose a detail level. Standard keeps the report smaller; High renders sharper page images and makes a larger file.
Leave sensitivity at Normal for two digital exports. Use Relaxed for two scans of the same paper, and Strict when you want the faintest change flagged.
Click "Compare PDFs" and download comparison.html, then open it and use the jump list to step through the changed pages.

Who it is for

Compare PDF is for the review round where you need to know what actually changed: a contract returned with amendments, a supplier's revised quotation, a designer's second draft, or two exports of the same report where one figure moved. It is faster and more reliable than reading both documents side by side, and it produces a record you can send to someone else as evidence of what differed.

How it works

Each page pair is compared twice. A block-based pixel pass outlines every region that changed, using an integer luma comparison with a per-block early exit.
A word-level difference then lists the text that used to be there against the text that replaced it, so the change is named rather than just located.
The report is one self-contained HTML file with the page images embedded, so it opens offline, prints, and can be emailed without the original PDFs.

What it cannot do

Two scans of the same page are never pixel-identical, so a strict comparison outlines the whole sheet. Relaxed sensitivity exists for exactly that case.
A page can show as changed with identical text, because something visible moved — a re-saved PDF shifts glyphs by a fraction of a point.
A hundred pages is the ceiling per run, and long documents render at a smaller scale to keep the report openable.
It compares page by page, so an inserted page shifts everything after it and shows as changed.

Frequently asked questions

What do the red outlines mean?

They mark regions where the rendered pixels differ. A moved logo, a changed figure in a table, an extra paragraph — all show as boxes on the page image. The word diff underneath tells you what actually changed in the text, which is more precise when the change is wording.

Why does a page show as changed when the text is identical?

Because something visible moved. A re-saved PDF can shift text by a fraction of a point, and a stamped date or a replaced image changes pixels without changing words. Look at the outlines: they show where.

What sensitivity should I use for scanned documents?

Relaxed. Two scans of the same sheet are never pixel-identical — paper texture, dust and scanner variation all differ — so a strict comparison outlines the whole page and tells you nothing. Relaxed tolerates that and flags only the substantial differences.

Does it work on documents with different page counts?

Yes. Pages that exist in only one file are marked as such, so an inserted or deleted page is obvious rather than being silently compared against the wrong page.

Are my documents uploaded?

No. Both PDFs are rendered and diffed in your browser and the report is generated there, so two versions of a confidential contract never leave your device.