Two versions of the same contract, invoice or report — and you need to know, precisely, what changed between them.
Upload the original and the changed file and get a page-by-page breakdown: every changed word named directly, plus any visual difference — a replaced logo, a moved signature, an added stamp — that a text-only comparison would miss entirely.
By ShinobiTools Team · Last updated: August 2026
This checks the words AND the pixels — not just one. A tool that only diffs the text layer can miss a replaced logo or a moved signature and still say "no differences" — that isn't an honest comparison. This tool runs both: a word-level text diff that names the changed word directly, and a rendered-pixel comparison on every page that catches visual changes text alone can't see. Small rendering noise — a point or two of drift from re-saving a file — is filtered out so a re-saved PDF doesn't come back "completely different". Your files are only ever handled in memory here, never written to our disk.
A PDF page is really two things at once: the text you can select and search, and the pixels you actually see. A tool that only diffs the text layer is blind to anything that isn't text — a swapped photo, a replaced logo, a signature block that moved a few centimetres. It will happily report "no differences" while the page looks completely different.
This tool renders every shared page and compares the pixels as well as the words, so a visual-only change — one that never touches the text layer — still gets caught and reported as its own kind of finding, separate from the word changes.
Re-saving, re-printing or re-scanning a PDF almost always shifts something by a fraction of a point — that's normal rendering drift, not a real change, and a comparison tool that flags it is a comparison tool nobody can trust. This one only reports a visual difference once the affected area of the page passes a threshold well above that drift: a rounding-level shift stays silent, a genuinely moved, resized or replaced element gets flagged.
Both PDFs are processed entirely in memory on our own machine and the comparison is handed back in the response — neither file is ever written to our disk for this tool, and nothing is kept afterwards.
Found a change you need gone for good, not just flagged? Redact PDF cuts marked content out of the file itself, not just paints over it. Comparing two contract drafts before combining them into one packet? Merge PDF puts multiple files together in the order you set. Found a change worth pointing out rather than acting on yet? Annotate PDF highlights it or drops a note right where it is. Want to ask specific questions about either version instead of scanning the whole diff yourself? Chat with PDF answers questions about a document's content directly.
Upload the original in the first box and the changed version in the second, then press Compare files. You get a page-by-page breakdown of what's different: changed words named directly, and any visual change flagged separately.
A text diff only reads the words in the file. It can't see a replaced logo, a swapped photo or a moved signature, because none of that is text. This tool renders each page too, so a "no differences" result means no differences, not just no changed words, even when two pages look nearly the same at a glance. The visual check is also what makes a compare in PDF pages with images trustworthy.
No. Tiny rendering differences of a point or two, the kind any PDF viewer introduces when it re-saves a file, are filtered out before something counts as a visual change. Only a genuine shift or replacement is reported.
When you compare PDF documents of different lengths, the pages both files share are compared normally, and any extra pages at the end of the longer file are listed separately rather than treated as an error.
No. When you compare PDF files here you get a page-by-page list with the changes named directly, plus a picture of the changed page wherever something visual moved, but no side-by-side view.
No. It tells you a page's appearance changed and shows a rendering of it so you can look, but it doesn't pinpoint the exact pixels that moved. To mark and permanently remove specific content, use Redact PDF instead.
Partly. A scanned page is a picture with no text layer, so there are no words to compare, but the visual comparison still runs and flags a page whose appearance changed. For a word-by-word check, both files need real text, like a PDF exported from Word.
Yes. Any two PDF files work, whatever created them; both the text and the rendered appearance are checked, so content and layout differences are both caught. A PDF doc from one program and a PDF file from another compare the same way.
If a formatting change alters the words or visibly changes how a page looks, such as a different font or size, it is reported. Invisible formatting metadata that does not affect the text or appearance is not.
25 pages and 250 MB per file on the free tier, or 200 pages and 500 MB with a pass. There is no account, no watermark and no daily cap.
Same idea, a page-by-page diff between two versions, but with no Acrobat licence or account: drop both files in the browser and get the report back directly. Acrobat's tool is good if you already pay for the subscription; this one is free.
ShinobiTools · KvK 42147182 · info@shinobitools.com