Compare PDF Files

What PDF comparison actually does 🔍

Comparing two PDFs means asking a simple question: does page 1 of document A match page 1 of document B? If not, what's different? That question turns out to be harder to answer than it looks, because PDFs are not text files — they're rendering recipes.

Two PDFs can have identical visible content but completely different bytes, because of how they were produced. Word exports look different from Chrome exports look different from LaTeX exports, even for the same content. Two versions of the same page might differ by a single digit or by an entire paragraph — and the visual difference may be obvious or invisible at a glance.

The reliable approach is visual comparison: render both pages to a canvas, compare them pixel by pixel, and highlight the regions that differ. That's what this tool does.

Three ways to compare 📊

Each view answers a slightly different question. Switch between them depending on what you're looking for.

How to compare two PDFs — step by step 📋

1. Load PDF A into the top upload slot — this is the original or reference version.
2. Load PDF B into the bottom upload slot — this is the revised version you want to check.
3. Wait for processing. Every page of both PDFs is rendered and compared automatically. Progress shows in the sidebar.
4. Review the summary. Identical, different, and only-in-one counts appear in the sidebar along with the overall similarity percentage.
5. Navigate to any page with the page list or the navigation buttons. Pages with the biggest differences are easy to spot — they show a red dot in the list.
6. Switch between view modes. Side by side for reading, overlay for layout shifts, diff for a quick highlight.
7. Adjust the threshold if needed. If a page looks identical but shows as different, increase the threshold. If a tiny change isn't being flagged, decrease it.

When to use this tool 📋

How the diff algorithm works 🔬

The tool follows the same pipeline that desktop diff software like Adobe Acrobat's "Compare Files" uses:

1. Render both pages to canvases at the same scale.
2. Read the pixel data for each canvas using getImageData().
3. Compare pixel by pixel — for each position, compute the average color difference between A and B.
4. Apply the threshold — if the difference exceeds your threshold (default 30 out of 255), the pixel is marked as changed.
5. Count changed pixels and compute similarity = 100 × (unchanged / total).
6. Render the diff by drawing A's pixels faded, then overdrawing changed pixels in red.

Why a browser-based comparator matters 🔒

Contract revisions, medical records, legal filings, confidential business proposals — these are precisely the documents where you'd want to compare versions, and precisely the documents you don't want to upload to a stranger's server.

This tool runs entirely in your browser. Both PDFs are read into memory locally, rendered to canvases on your CPU, and compared pixel by pixel without any network activity. Close the tab and every trace is gone.

• No server uploads — neither PDF leaves your device.
• No sign-up — no email, no account, no waiting.
• No page limits — only your browser's memory constrains you.
• Works offline — after the page loads, turn off Wi-Fi and it still works.
• No watermarks — the tool never modifies your PDFs.
• Free forever — no ads inside the tool, no per-file fee.

Who uses a PDF comparer? 👥

• Lawyers — checking that a redlined contract matches what was agreed.
• Editors and proofreaders — confirming that only marked changes were made.
• Compliance officers — auditing document revisions against policy.
• Project managers — verifying a deliverable matches the specification.
• Designers — spotting layout shifts between proof versions.
• Engineers — comparing exported drawings or reports across revisions.
• Anyone who's received two versions of a document and needs to know what changed.

Common problems and how to fix them 🛠️

"Every page shows as different, but they look the same"

Almost certainly a threshold issue. Two PDFs of the same content exported from different tools will have pixel differences of 5–15 due to font hinting and anti-aliasing. Increase the threshold to 50 or 60 — the visible differences will still be flagged, but the invisible noise disappears.

"A tiny change isn't being detected"

Lower the threshold to 15 or 20. Small text changes (a single digit or a comma) can produce pixel differences of only 20–30 in a small region. The lower threshold catches these but may introduce false positives elsewhere.

"PDFs with different page sizes show as different everywhere"

That's expected. If page A is A4 and page B is Letter, the canvas sizes differ and the comparison treats the smaller page's overflow area as changed. The tool uses the larger of the two dimensions and fills the missing regions with white. For an accurate comparison, resize one PDF to match the other with our PDF Page Resizer tool first.

"Scanned PDFs show as different even though they look the same"

Scans vary slightly between captures — lighting, dust, paper texture, and sensor noise all produce pixel-level differences. Set the threshold to 70 or higher. Real content changes will still be caught, but the noise won't.

"The page takes a long time to load"

Very large PDFs (100+ pages each) can take 30–60 seconds to process at default settings. The tool yields to the UI between pages so the progress bar updates, but rendering is CPU-bound. For very large documents, consider splitting them first with our Split PDF tool and comparing section by section.

"I only want to see the pages that changed"

Use the page list in the sidebar. Every page shows a colored dot: green for identical, amber for minor differences below the "different" threshold, red for pages with substantial changes. Click any page to jump to it in the comparison view.

Is it safe to compare PDFs here? 🔐

Yes. Here's exactly what happens when you use this tool:

1. You pick two PDFs from your computer.
2. Both are read into memory locally.
3. Pages are rendered to canvases on your device.
4. Pixel data is compared in JavaScript — nothing leaves the tab.
5. When you close the tab, every trace disappears.

At no point does either PDF leave your device. No data is sent to SkillsInsider's servers. No tracking cookies watch your comparison. Turn off Wi-Fi after the page loads and it still works perfectly.