PDF Compare

Align pages and inspect text changes alongside actual visual differences.

Processed on your deviceFree · No account needed

Compare extracted text and rendered page images separately. Review the proposed page mapping before running the comparison.

Before / left PDF

After / right PDF

Per PDF: 10 MiB, 1–20 pages, 200,000 text characters. Everything is processed locally. Original files remain unchanged.

Example files and comparison limits

The examples include a one-cent amount change, an inserted page, identical text with different wrapping and position, ambiguous repeated pages and a changed image-only page.

Text uses the PDF’s extracted reading order. Spaces and line wrapping are ignored; punctuation, hyphens and numbers remain significant. Columns and unusual font encodings can affect extraction. Content moved across page boundaries may appear as removals and additions. This tool does not perform OCR. Scan pages only receive a visual comparison, even when images appear unchanged.

Images use a common physical scale with at most 900 px width. Rendering, anti-aliasing and the selected threshold affect differences; very small details may be missed. Up to 48 MP are rendered for the comparison. Text is limited to 30,000 tokens per PDF and 8,000 per page. Complex text changes are marked unresolved. Metadata, attachments, interactive form behavior and signature validity are not compared. No legal assessment or proof of document equivalence.

Preparing your tool…

How to use it

Choose the before and after PDFs or load the example pair, then read their text and page previews.
Review suggested page pairs, ambiguous alternatives and pages without a counterpart. Correct the mapping and explicitly confirm it.
Choose case sensitivity and the pixel threshold, then compare extracted text and rendered images separately.
Inspect changes at their approximate original locations and download the JSON/XLSX report or the selected difference PNG.

The example changes EUR 1000.00 to EUR 1000.01, inserts a page and moves unchanged text. Repeated annexes remain ambiguous; the image-only final page receives a visual comparison.

Details & limits

Two PDFs up to 10 MiB and 20 pages each. Confirm ambiguous page matches. Scanned pages receive visual comparison, not semantic interpretation. Review source regions and all unmatched pages.

Does an unchanged result prove that two PDFs are equivalent?

No. Text comparison ignores whitespace and line wrapping within a page pair; cross-page moves may appear as removals/additions. Visual comparison uses a common scale, at most 900 px width and an explicit RGB threshold, so small details can be missed. Scans receive no semantic text comparison or OCR. Complex text differences remain unresolved. Metadata, attachments, interactive forms and signature validity are outside the comparison. Up to 10 MiB/20 pages per PDF. There is no legal assessment; only the separate full-byte check can report byte-for-byte identity.

PDF Compare

How to use it

Details & limits

Related tools