About DocDiffer
DocDiffer compares two versions of a document and shows what changed between them — which paragraphs were added, which were removed, and which words were rewritten inside the ones that stayed. It was built for the moment a new draft lands in your inbox with a note saying “only minor updates”, and you need an answer that does not depend on the note.
It runs entirely in your browser. Both files are opened by the browser itself, their text is rebuilt locally, and the comparison is JavaScript running in this tab — there is no upload, no account, no analytics and no server that could keep a copy. Once the page has loaded it works offline, which is the simplest proof that the documents are not going anywhere.
It is built and maintained by Interhive OÜ, a small studio in Tallinn that makes tiny, dependable web tools.
What it does
- Paragraph alignment
Both documents are reduced to paragraphs and aligned with a Myers diff, so a clause that stayed put stays matched with itself even after everything around it moved.
- Word-level marks
Paragraphs that were rewritten rather than replaced are diffed again inside, and the removed and inserted words are struck through and underlined in place. You can switch the detail down to sentence or paragraph.
- Five formats, mixed freely
PDF, DOCX, TXT, Markdown and RTF, plus text pasted straight into either side. Extraction reports what it found — pages, paragraphs, words — and warns when a PDF has no text layer.
- A report you can keep
Every change carries a page and paragraph reference. The full list downloads as plain text, Markdown or a self-contained HTML page that prints as a readable redline.
What it does not do
It compares text, not formatting: fonts, colours, margins, styles, comments and tracked-change marks are left out on purpose. It cannot read a scanned PDF that has no text layer, or legacy .doc, Pages and ODT files — export those to DOCX or PDF first. And it does not try to tell you which of the changes it found matter; that judgement is yours.
Get in touch
Found a document that extracts badly, or want a format supported? Email us — we read everything.
support@interhive.org