Skip to main content
OS Authority

Compare the text in two PDF files

Find text added or removed between two versions of a PDF. Review corresponding pages by word or line and download a plain-text report. Both documents stay on your device.

Your PDFs and extracted text stay on this device. We do not upload or log document contents.

How it works

  1. Choose the original PDF, then the new version. Reorder them if necessary.
  2. Choose words or lines and decide whether to ignore repeated whitespace.
  3. Run the comparison and review each page’s status and labeled changes.
  4. Download the text report. Check any unprocessed pages and visual details separately.

A page-by-page text comparison

Page 1 is compared with page 1, page 2 with page 2, and so on. A changed total at the end is reported as a page present in only one document. An inserted page in the middle can make later page pairs differ; this version does not automatically align moved pages. For a two-page draft, replacing one word on page 2 produces changes on that page while page 1’s extracted text can match.

Words, lines and whitespace

Words mode highlights token changes; lines mode compares whole extracted lines. Repeated-whitespace normalization can suppress differences caused by runs of spaces or tabs, while line boundaries remain meaningful in lines mode. PDF extraction order depends on the document: columns, unusual fonts and layout changes can produce different extracted text even when a page looks similar.

What this comparison cannot establish

Images, scanned text without a text layer, page layout, font styling, annotations, form values and signatures are outside this text-only comparison. Pages without extractable text are marked separately, not declared identical. Matching extracted text is not proof that two PDFs have the same appearance, contents or validity. This tool does not perform OCR or a visual pixel comparison.

Limits and incomplete reports

Each input is limited to 20 MB. Combined inputs are limited to 20 MB on phones or low-memory devices, or 40 MB on larger devices; page caps are 100 or 200 per document. Extracted text is capped at 250,000 or 500,000 characters across both PDFs, with 20,000 characters per page. A page comparison allows 2,000 combined tokens, retained results allow 5,000 diff parts, and the report is capped at 2 MB. A limit failure is marked incomplete; counts cover only processed text.

Frequently asked questions

Can it compare scanned PDFs?

Only text already present in a PDF text layer can be extracted. Image-only pages are marked as having no extractable text. Use an OCR workflow separately if you need scanned words compared.

Does matching text mean the files are identical?

No. Pictures, page layout, signatures and other PDF features can differ while extracted text matches. Review those separately.

How are added and removed words counted?

Added means text in the new version; removed means text in the original. Whitespace-only tokens do not count as words. Swapping the documents reverses the direction of changes.

Can I compare a password-protected PDF?

Unlock it first using a password you are authorized to use. The comparison tool does not bypass encryption.

Are my documents uploaded?

No. Extraction and comparison run in a browser worker, and the report is created locally.

Last reviewed: October 10, 2026

Related tools