PDF Diff Checker

Compare two PDF files side by side. Changed regions light up in red (removed) and green (added). Every page is rendered and compared locally — your files never leave your browser.

📄
VERSION 1 (Old)
Drop PDF or click to browse
📄
VERSION 2 (New)
Drop PDF or click to browse
Pages compared
Pages changed
Pages identical
Page count diff
Removed / changed pixels
Added / changed pixels
Identical pixels

How it works

All comparison logic runs entirely in your browser using two open-source libraries — no server, no account, no file upload required.

1. Render with PDF.js Each page of both PDFs is drawn onto a hidden HTML canvas at 150 dpi using Mozilla's PDF.js engine. This captures text, images, and vector graphics identically to a real PDF viewer.
2. Pixel comparison with pixelmatch The raw pixel arrays from matching pages are fed into pixelmatch, which compares every RGBA pixel and writes a diff image — red for pixels only in V1, green for pixels only in V2, and transparent for identical pixels.
3. Change-density score The number of mismatched pixels divided by the total page area gives a change-density percentage (0 % = identical, 100 % = completely different). This drives the summary bars shown above the page panels.
4. Page count mismatch If V1 and V2 have different page counts, the extra pages in the longer file are shown individually without a diff partner. They are flagged as "only in V1" or "only in V2" in the summary.

Frequently asked questions

Are my PDF files uploaded to a server?
No — never. The tool uses the browser's built-in File API and processes everything locally with JavaScript. The files are read into memory on your device and are never sent over the network. You can verify this by opening your browser's DevTools → Network tab and confirming there are no outgoing requests when you click Compare.
What does the diff overlay actually show?
The middle "Diff" column shows a pixel-level overlay on top of V1. Pixels that are present in V1 but not V2 are tinted red; pixels that are new in V2 appear green. Unchanged pixels are rendered at reduced opacity so you can still read the background. A change-density of 0 % means every pixel is identical; anything above ~1 % usually indicates a visible content change.
What are the limitations of this pixel-based approach?
Because the comparison is visual rather than text-extraction based, minor rendering differences (sub-pixel font hinting, embedded image compression artefacts) can show up as small amounts of noise even when the text content is unchanged. If two PDFs were produced by different PDF generators or exported at different quality settings, you may see a low-level background noise of 0.5 – 2 %. For contract or legal review where text changes matter most, check the diff overlay at high zoom against any pages flagged as changed.
Can I compare PDFs with different page counts?
Yes. The tool compares as many pages as the shorter file has. Any extra pages in the longer file are shown in a separate section labelled "only in V1" or "only in V2". The summary card shows the page count difference so you always know at a glance whether one document has grown or shrunk.
What is the change-density percentage in the bar chart?
It is the fraction of pixels on that page that differ between V1 and V2, expressed as a percentage of the page's total pixel area. A value near 0 % usually means only minor typographic rounding; above about 3 % usually corresponds to a visible change such as a new paragraph, an image swap, or a relocated block of text.