textdiff

textdiff / Compare PDF

Compare PDF files: find the differences

Open two PDFs and see every changed word, number and clause highlighted. The text is extracted on your device, so confidential documents never leave it.

0 chars · 0 lines
0 chars · 0 lines

What happens when you open a PDF

  1. PDF.js reads each page's text layer, in reading order, inside your browser tab.
  2. The two extracted texts are lined up and compared with the Myers diff algorithm, then each changed line is compared word by word.
  3. You get the changes side by side (or inline on a phone), a count of added and removed words, and a strip down the edge showing where in the document each change sits.

Getting a clean comparison

  • Keep "Ignore whitespace" on. PDFs store text as positioned lines, so a sentence that wraps differently would otherwise show as changed.
  • Use word mode for prose and character mode for figures. Character mode catches a single changed digit in an amount or date.
  • Headers and footers repeat on every page. If page numbers shifted, expect small changes at page boundaries.
  • Scanned PDFs won't work: there's no text layer to read. Comparing the original Word files with document compare is often easier.

Questions

How do I compare two PDF files for differences?

Click Open file on each side and choose the two PDFs, or drag both onto the page at once. The text of every page is extracted in your browser and compared word by word, and the differences are highlighted side by side or inline.

Are my PDFs uploaded?

No. Text is extracted on your device with Mozilla's open-source PDF.js and compared in your browser. Only if you click "Summarise the changes with AI" are the changed passages (not the whole file) sent to generate a summary.

Why does it say no text was found?

The PDF is probably a scan: each page is a picture of text rather than text. Comparing scans needs OCR (optical character recognition), which this tool does not do. Run the scans through an OCR tool first, or compare the original Word or text files.

Does it compare images, layout or formatting?

No, it compares the words. Fonts, colours, images, page layout and annotations are ignored, which keeps the result focused on wording changes. Re-flowed line breaks can look like changes, so "Ignore whitespace" is on by default here.

Can I compare a PDF with a Word document?

Yes. Open the PDF on one side and the .docx on the other. Line breaks will differ between the two formats, but with "Ignore whitespace" on and word mode, wording changes stand out clearly.

Is there a page limit?

There is no fixed page count. Files up to 20 MB and up to 1,000,000 characters of extracted text per side are supported; a 100-page contract is typically well under that.