Compare PDFs
Upload two PDF files and get a line-by-line comparison of their embedded text, with additions and deletions clearly highlighted. This is a real text comparison, not OCR, and the files are processed entirely on your device.
Your two files are processed on your device and are never uploaded to our servers.
This tool compares the text that already exists inside each file (it's not OCR), and shows the lines added and removed between them.
First file
Drag a PDF file or choose one
Second file
Drag a PDF file or choose one
What this tool does
Compares the text of two PDFs and shows where they differ, line by line, along with each one's page count. The comparison is on the text layer stored in the files, not on how the pages look.
When to use it
- You received a revised contract and need to know exactly what changed.
- You are checking whether two seemingly identical files really are, before relying on one.
- You are reviewing two successive drafts of a report and want the edits isolated.
How to use it
- Put the first document in the first file slot.
- Put the second document in the second file slot.
- Run the comparison and read the result: either the files are textually identical, or you get the differing lines.
A practical example
Input
Two versions of a contract, the second with one clause edited and a paragraph added.
Output
A result showing any page-count difference, the edited clause's line, and the added paragraph — while every matching line goes unmentioned.
Limitations
- The comparison is purely textual: it does not detect changes to fonts, colours, element positions or images. Two files with identical text and completely different design will be reported as matching.
- It does not work on scanned documents, which carry no text layer — the same constraint as the PDF to Text tool.
- Multi-column documents can produce interleaved text and therefore differences that are not real.
Privacy and processing
Both files are read and compared on your device; neither is uploaded and their text is never sent anywhere.
Technical notes
- The comparison uses a line-level longest-common-subsequence diff — the same algorithm as the plain-text diff tool, so the two behave consistently.