A changed figure is easy to miss
A revised quotation may change a price in one line. A new report may move an entire section. Compare PDF gives you separate text and visual evidence so you can check both without switching between browser tabs.
Find changes in text, images and page layout. Open the original and revised PDFs, review them side by side, and keep a report of what changed.
The earlier version you want to check against.
The newer version, with any updates or additions.
Smart align uses selectable text when available and a low-resolution visual fingerprint as supporting evidence. It is useful when pages are inserted or scans shift the numbering. Always review the proposed pairs.
Use a dash for a missing page. Pages omitted from this list are not assessed. Each page can appear once on its side.
Yes. Visual comparison works for scanned pages. OCR Assist can optionally recognize scan text for matching and text differences, but OCR errors are possible and should be checked against the visual result.
OCR Assist downloads the recognition runtime and selected language data only when needed. It can improve scan text comparison, but OCR mistakes can also appear as differences.
Your original files are never changed. A comparison is not a redaction, signature check or guarantee that two documents mean the same thing.
This summary compares exposed PDF metadata, page count, page labels, outline/bookmark count, attachments and form-field counts where the browser PDF library can read them. It does not validate signatures or hidden object equivalence.
Rendering page pair…
Hold Space and drag to pan. Use ← / → for page pairs. Text fields keep normal keyboard behaviour. Highlight boxes are only guides; they do not remove information.
Reports contain file names and findings. The visual PDF and PNG also contain original page content. They are not redacted or encrypted. Keep them private.
A revised quotation may change a price in one line. A new report may move an entire section. Compare PDF gives you separate text and visual evidence so you can check both without switching between browser tabs.
An inserted page shifts later page numbers. Smart align can use both selectable text and page appearance to propose page pairs across that shift. Always inspect the pair list, particularly for scans, repeated pages and substantially rewritten documents.
Download the settings, coverage and findings with your comparison report. Share it with the right reviewer rather than treating a no-difference result as proof that the documents have the same legal or practical meaning.
Choose the older PDF as A and the revised PDF as B. Start with text and visual appearance, then select how pages should be paired. Press Compare both PDFs. The report separates changed pages, pages with no differences detected, and pages that could not be assessed. Select a page pair to review the actual previews and extracted text changes.
Use the wipe view to reveal one version over the other. The side-by-side view keeps the files separate and can synchronise scrolling. When you have checked the page pairs, download a visual PDF report or export the findings as JSON, CSV or plain text. Neither original file is edited.
Text comparison looks at the selectable text extracted from the PDF. It can show added or removed words, numbers and punctuation. Whitespace is ignored by default, and you can choose whether letter case matters. Font encoding, reading order, hidden OCR text and unusual layouts can affect extraction. A scanned page without a text layer is not assessed in text-only mode.
Visual comparison renders pages onto a common canvas at the chosen resolution. It checks pixel colour differences above your tolerance and groups nearby changes into approximate boxes. A shifted paragraph, a changed image or a different signature appearance may show up visually even when extracted text is unchanged. Those boxes are not precise semantic edits, and no automatic signature verification is performed.
Use visual appearance for image-only scans. If text matters, optionally enable OCR Assist so image-only pages can contribute recognized text to page matching and text comparison. OCR can misread characters, so visual evidence remains important. Small scan or render shifts can be compensated automatically, and the revised PDF can be rotated for comparison without changing the original file.
At 144 DPI, finer differences are sampled than at 96 DPI, with more memory and processing work. Increasing colour tolerance may reduce mild scan noise but can also hide light-coloured edits. The report records the setting so that a reviewer can understand the scope of the result.
Optional top and bottom margins can exclude a timestamp or page number from comparison. Enter the margins in PDF points: 72 points equal one inch. These margins affect visual comparison and text whose baseline is inside the ignored area. Do not exclude a header containing a meaningful revision date or reference number unless that is intentional. Exclusions appear in downloaded reports.
Page-order pairing is the clearest starting point for files with the same structure. Smart alignment scores text and a small visual fingerprint while keeping page order; exact-text alignment remains available when you want conservative unchanged-text anchors. Manual pairing allows a selected comparison such as 1:1, 2:- and -:3. A dash means a page is absent on that side. Pages you leave out of a manual list are explicitly outside the review.
The documents are processed in your browser, without an account or a conversion service. PDF libraries are hosted with UVs Tool Hub and load after you choose a document. This tool does not save an automatic recovery copy. Reloading or clearing the session removes the current comparison; downloaded reports remain wherever you save them.
Each file can contain up to 60 pages and be up to 50 MB. There are also rendering and total-pixel limits to reduce the risk of exhausting browser memory. A large-format drawing can reach those limits before a normal A4 document. Lower the comparison resolution or use Split PDF to prepare a smaller part. For private content that must be removed before sharing, use PDF Redactor; comparison highlighting is not redaction.
Yes. The text comparison, visual previews and included report downloads are free. No account or conversion credits are required. Browser resource limits still apply.
The PDFs are processed locally in your browser. The tool loads its libraries from UVs Tool Hub, but does not upload document content for comparison. Reports are downloaded only when you choose an export action.
Yes. Visual comparison works for scanned pages. OCR Assist can optionally recognize scan text for matching and text differences, but OCR errors are possible and should be checked against the visual result.
No. It means none was detected in the selected pages and modes at the chosen resolution, tolerance and exclusions. Metadata, attachments, hidden content, signature validity and meaning are not compared. Matching original SHA-256 fingerprints are reported separately when available.
Page-order mode shows unpaired extra pages as added or removed. Exact-text alignment can use unchanged text pages as anchors. Manual pairing lets you review a selected set. The report shows how many source pages are covered.
Yes. Download a visual PDF report, a PNG of the current page pair, or JSON, CSV and plain-text findings. Changed text is included in data reports only when you enable that option. Visual reports contain original page content and should be kept private.