How to compare two PDF documents
- Choose two PDFs. Select one or two files in the upload area. If you select one, its first page appears immediately and you can add the second PDF inside the workspace. Enter a password if requested.
- Choose how to compare. Use text, visual, or combined comparison. Optionally select page ranges and adjust sensitivity.
- Review the differences. Check the suggested page pairs. Select a change to highlight its location in the previews.
- Keep a report. Download a standalone HTML report with the settings, page pairs, thumbnails, and change descriptions. Your original PDFs are not modified.
Keep the earlier document in Original and the newer one in Revised. This direction determines whether wording is labelled added or removed. If the files are reversed, use Swap before running the comparison. You can replace either file without choosing the other one again.
Text comparison or visual comparison?
Text comparison looks for additions and removals in extracted wording. It is useful for checking revised instructions, product descriptions, reference numbers, or amounts. It does not understand the meaning of a change. Replacing one word may appear as a removal followed by an addition rather than a single replacement.
Visual comparison looks at rendered page images. It can reveal changes to logos, diagrams, colours, spacing, and the visible appearance of text. A highlighted visual region means pixels differ at the comparison resolution and sensitivity. It is not a claim that every pixel represents a meaningful edit.
Choose Text and visual when both wording and presentation matter. Use Text only when formatting is intentionally different. Use Visual only for scans, drawings, or image-heavy pages. Neither mode examines every internal object stored inside a PDF.
Understand the difference highlights
Only in first PDF means wording found in the first document but not the second. Only in second PDF means wording found in the second document but not the first. These descriptions replace the technical labels removed and added text. No wording is actually deleted or inserted in either file. Visual differences highlight changed appearance.
Each preview has floating page arrows, a page number input, zoom controls, and Fit width. Desktop previews start at 75%; mobile previews initially fit the available width. After comparison, page arrows follow matched page pairs so the comparison stays aligned. Use the separate Previous difference and Next difference buttons to move through individual changes. Show settings & results opens the full changes list; hiding it expands the previews.
Text highlights use positions from the PDF text layer. Individual word positions are estimated within text runs. Rotated text, complex columns, unusual fonts, and tables can make the highlight less precise or alter extraction order. Always read the surrounding content before deciding whether a change is significant.
Review page matching before interpreting results
An extra cover sheet can shift every following page number. Automatic matching considers page wording and a small visual fingerprint to suggest corresponding pages while preserving sequence. Unmatched pages are shown as possible additions or removals. Scan-based and low-confidence matches are marked for review.
Repeated forms, nearly blank pages, and heavily revised layouts can confuse matching. For a known pair, enter its page number in each range and choose Selected page order. For example, compare original page 3 with revised page 4 directly. Matching is not automatic detection of every moved page, and it does not reconstruct paragraphs that reflow across several pages.
Compare scans without mistaking noise for edits
Two scans of the same paper can differ because the sheet moved slightly, the scanner changed exposure, or an image was compressed. These changes may generate large highlighted regions. Start with straight, readable scans and review the Balanced setting before trying a lower visual sensitivity.
This tool does not run OCR automatically. If a scan contains no selectable text, text comparison cannot tell you whether its words changed. Use visual results, or create searchable copies with OCR PDF before comparing wording. OCR can introduce recognition mistakes, so check those results against the scans too.
Choose settings that match your review
Ignore letter case when capitalization is unimportant. Ignore whitespace when differences in spaces or line breaks would distract from wording. Leave these options off when those details matter. High visual sensitivity detects smaller colour differences but can produce more noise; Low sensitivity can miss subtle changes.
Changing files or comparison settings invalidates the current results. Run the comparison again before downloading a report. The report records the settings and selected pages, so someone reviewing it can understand what was checked and what was outside the comparison.
Useful document review workflows
For an updated instruction manual, check text changes first, then inspect altered diagrams visually. For an invoice revision, review numbers and descriptions alongside page images because extraction order may not follow table rows. For a revised brochure, visual differences can reveal moved logos, changed colours, or altered spacing.
Comparison assists review; it does not replace checking important documents. A summary of changes cannot authenticate a file, validate a digital signature, determine whether a revision is authorized, or establish that an agreement has the same meaning.
Private processing and practical limits
PDF content stays on your device. The comparison libraries load when the tool is used, and images are processed page by page to reduce memory use. There are no automatic document drafts or cloud comparison jobs. Downloaded reports may contain sensitive excerpts and thumbnails, so store and share them as carefully as the source files.
A modern desktop browser generally has more room for complex PDFs than a phone. If processing is slow, compare a smaller page range or use Text only for documents with reliable text layers. Visual comparison uses a bounded rendering resolution, so extremely fine details may not be detected. Cancelled or failed comparisons are not reported as complete.
Frequently asked questions
Are my PDFs uploaded to a server?
No. Opening, rendering, comparison, and report generation happen in your browser. The tool does not save document drafts or send their contents to a comparison service.
Can I compare scanned PDFs?
Yes, using visual or combined comparison. The tool compares rendered page images. It does not run OCR, so scans without a text layer cannot produce word-level results. A notice identifies pages where text is unavailable.
Why do similar-looking scans show differences?
Scanner noise, skew, compression, and small shifts can change pixels even when the wording is unchanged. Try a lower sensitivity and review the highlighted regions. Automatic deskew and image registration are not included.
Does no difference mean the files are identical?
No. Results apply to the selected pages, mode, and sensitivity. Hidden metadata, attachments, scripts, and digital signature validity are not compared. Text-only results cannot establish that page images or formatting are unchanged.
What happens if a page was inserted?
Automatic matching suggests corresponding pages in sequence and can identify unmatched additions or removals. Repeated pages and scans may be ambiguous. Review the pairs and use explicit ranges with selected page order when you need to correct a match.
Can I compare password-protected PDFs?
Yes, when they can be opened with a password you know. The shared password prompt unlocks a local working copy. Original files remain unchanged. Only compare documents you are authorized to open.
Can I compare a long document?
Select up to 200 pages per PDF for each run, with files below 100 MB each. Large pages, complex graphics, and device memory may still limit processing. Use smaller ranges on a phone. Cancel stops the current run and does not produce a complete report.
Can I save the comparison as a PDF?
The download is a standalone HTML report. Open it in your browser and use Print, then Save as PDF if available. The report is a separate summary, not an edited or annotated copy of either source PDF.