Compare two PDFs locally with word changes, page overlays and pixel highlights. Review page-by-page text differences and download a comparison summary.
Check both extracted text and rendered pages; neither alone establishes complete document equality.
The common words stay unchanged. Use the page overlay to inspect styling or graphical changes that the extracted text cannot describe.
Pages are paired by page number. Added pages can shift comparisons; scanned pages may not contain searchable text.
Compare two PDF versions by loading an original and a revision, then selecting Compare PDFs. The tool extracts text page by page and renders paired pages for visual review. Added words are underlined in green and removed words are struck through in red. A blend overlay and pixel-difference view help inspect layout changes.
Pages are paired by their page number rather than matched automatically by content. Inserting a page near the beginning can shift the apparent differences on every later page. The text overview explicitly distinguishes matching extracted text, changed text, added pages and removed pages. A matching text result does not establish that diagrams, annotations, font styling or signatures are unchanged.
This browser-based comparison accepts two PDFs with up to one hundred pages and fifty megabytes each. Scanned pages may contain no extractable text, so the visual review remains important. Pixel highlights are sensitive to rendering and position differences; they are not a legal document verification or a signature-validation service. Download the JSON summary to retain the per-page text status, and keep both source documents for a complete review.
Choose two versions you are authorized to inspect. Password-protected documents need an unprotected copy; the tool does not guess or remove passwords.
Press Compare PDFs. The reader loads locally and processes each document in order. Longer files show page-level progress before the comparison becomes available.
Use Previous page, Next page or the page selector. Read the word changes, move the blend slider, or choose pixel highlights. Blank extracted text requires visual review.
Download the JSON text report for a page-by-page record. The report contains filenames and text statuses rather than a full annotated PDF or a signature certificate.
Uses bounded sequence comparison for words, switching to coarser paragraphs on long pages to limit memory growth.
Displays the original, revised and blended or highlighted page at a shared scale and canvas size.
Summarizes extracted-text equality and missing pages without treating matching text as proof of complete document equality.
Locate wording changes between draft versions, then inspect charts and page layout in the overlays.
Check whether a refreshed report changed text, added pages or moved graphical elements before distribution.
Use the rendered comparison when OCR text is absent, while recognizing that scan alignment can create many pixel differences.
Load the original and revised PDFs and click Compare PDFs. Review the text highlights and rendered page comparison, moving through every relevant page.
It renders scanned pages for visual comparison but does not perform OCR. A page with no text layer can have an empty text result even when the images differ.
No. It pairs page one with page one and so on. A page insertion shifts later comparisons, so review those pages manually.
They indicate rendered color differences above a threshold. A shifted page, changed font rendering or scan alignment can produce highlights even when the wording matches.
No. It is a review aid for text and page appearance. It does not verify signatures, certify authenticity or replace a complete document review.
The JSON summary records the two filenames and each paired page’s extracted-text status. It does not embed the documents or every highlighted word.
Drop files here or click to browse — processing stays on your device
Drop files here or click to browse — processing stays on your device
50 MB and 100 pages per PDF. Text comparison ignores repeated whitespace; visual overlays compare rendered pages. Scanned pages need visual review—OCR is not performed. Password-protected files need an unprotected copy.
