Review & data · guide
What Compare PDF files does
Compare PDF files finds what changed between two drafts, word by word, and gives you a Markdown report of what was added and removed. Pages are matched by their content before they are compared, so one inserted page does not make every page after it look rewritten. It compares the words, so a change of layout or formatting alone does not show up. Both files stay in your browser.
How to compare two PDF files
- Choose the two drafts Select both PDFs at once, the earlier version and the later one. Open Selected files to check that the earlier version is listed first. Neither file is uploaded.
- Compare them Select Compare. The pages of the two drafts are matched by content, then compared word by word.
- Read the changes Save the report, a Markdown file of what was added and removed, and check the changes that matter against the PDFs.
Worked example
Find what changed between version 1 and version 2 of a 12-page contract.
- Starting point
- Select contract-v1.pdf and contract-v2.pdf together, then open Selected files and check that contract-v1.pdf is listed first. Both need selectable text.
- What to do
- Select Compare, then save contract-v1-changes.md and open it in a text editor or a Markdown viewer.
- Result
- Expect a report headed What changed that names both files, counts the words added and the words removed, and says how many pages differ, such as 2 of 12 pages differ. Each changed page gets its own heading, such as ## Page 4. Added words follow **+**, and removed words follow **−** between ~~ marks, which a Markdown viewer shows struck through. Open both PDFs at page 4 to read the change in place.
If the result is not right
A PDF that asks for a password when it opens cannot be compared until you remove the password with Unlock PDF. If changes appear missing, test text selection and OCR image-only drafts where needed; form field answers and comments are not compared at all. Review layout, images and signatures visually as a separate check.
Limits worth knowing
- A changed picture or layout can be important even when extracted wording is identical. Scans may have no usable text layer.
Questions and answers
Does Compare PDF spot formatting changes?
No, it compares the words. Text that was added, removed or changed is reported, but a change of font, colour or layout with the same words is not. For a visual check, look at the two versions side by side.
What if pages were added or removed between drafts?
That is handled: pages are matched by their content before they are compared, keeping their order, and the result says how many pages were added in the second document or are missing from it, instead of every later page looking different. Two pages can be paired only when the words they share make up at least 60% of the words on both pages together; each pair is then compared word by word. A page rewritten so heavily that it falls below that is reported as one page removed and one added.
Can I compare two PDFs to see if they match?
Yes, for the words in the page text. When every page has the same words in the same order, the result says No wording changed; otherwise it counts the words added and removed and the pages they are on, and the Markdown report lists each change. Answers typed into form fields and comments are not read, so a changed answer or a new comment is not reported. Two files with the same words can still differ in images, fonts, layout or hidden properties, which this does not check. A scanned PDF has no words to compare until it has been through OCR.
Why does it report changes nobody made?
Because it compares words exactly as each PDF stores them. Punctuation counts as part of a word, so 2024 and 2024. differ, and so do capitals. A word hyphenated across a line break in one version, columns or text boxes stored in a different order, or one file that went through OCR will all show up as edits. So will an edit that pushes words over a page break: they are reported as removed at the end of one page and added at the top of the next. A paragraph inserted early on, or the same text exported at another paper size, can show changes on every later page and report pages as added or removed. Read a change at the top or bottom of a page as possibly moved, not edited. Check small changes against the PDFs, and where you can, compare two files exported from the same app in the same way.
Which file counts as the earlier version?
The first one in the Selected files list. Words only in it are reported as removed, and words only in the second file as added. Select both PDFs in one go, because choosing again replaces the first selection; if the list is the wrong way round, choose them again, or read the report the other way about.
What does the report look like?
A Markdown file, which is plain text: it opens in any text editor and reads more neatly in a Markdown viewer. By default it is named after the first PDF with -changes.md on the end, and opens with the heading What changed, the two file names and three counts: words added, words removed, and how many pages differ, such as 2 of 9 pages differ. Each changed page then has its own heading, such as Page 3, Page 3 → 4 when the page numbers differ, Page 5 — added or Page 7 — removed for a page found in only one file, with additions marked + and removals marked − and struck through. There is no marked-up PDF or side-by-side view, so open the two PDFs at those pages to see each change in place.