Question 1

Is my data sent anywhere?

Accepted Answer

No. PDF comparison runs 100% in your browser using JavaScript and pdf-lib. Your files never leave your device, and the tool works offline once the page is loaded.

Question 2

What does PDF Compare detect?

Accepted Answer

It compares page counts, per-page text content (line-by-line diff with a Levenshtein similarity score), metadata fields (title, author, subject, keywords, creator, producer, dates), structure (page sizes, embedded font names), and per-page image counts. Choose from 4 comparison modes: text-only, metadata-only, full, or structure.

Question 3

What extra features?

Accepted Answer

18 extras: page-range parser; text comparator (line-by-line diff); whitespace + case normalizers; metadata comparator; structure comparator (page count, sizes, fonts); unified-diff formatter; Levenshtein similarity score; page-by-page diff generator; diff statistics (added / removed / modified / unchanged lines); 4 multi-format renderers (text, HTML, CSV, JSON); copy + download; localStorage history (max 20); shareable URL; summary stats; page-count diff; metadata diff; font-usage diff; image-count diff; and a significant-changes finder that surfaces the pages with the most differences.

Question 4

How is the similarity percentage calculated?

Accepted Answer

Per page, the tool computes a Levenshtein edit distance between the two text layers (after optional whitespace / case normalization), then converts it to a 0–100% similarity score: 100 means identical, 0 means completely different. The overall document similarity is the average of all compared pages.

Question 5

Can I compare scanned PDFs?

Accepted Answer

Text comparison works on PDFs that contain an embedded text layer. Scanned PDFs without a text layer will report empty text on each page — structure and metadata comparisons still work. For true text comparison of scans, run an OCR pass first (see our PDF OCR Text Extractor) and then compare the OCR'd text.

PDF Compare

How to use