KenjaPDF is a minimal PDF reader with on-device translation for Mac. Drop in a document, tap once, and read the original side by side with its translation — scrolling in sync, layout preserved, nothing uploaded.
From foreign PDF to bilingual reading without leaving the app — read it, translate it, export it.
Drag a PDF onto the window, or press ⌘O. Continuous scrolling, thumbnails and outline in a collapsible sidebar, free zoom, page labels, printing — and password-protected files unlock with a prompt.
Pick a language pair — the source is auto-detected — and press Translate. Apple's on-device engine translates the whole document, and the result opens beside the original, scrolling in sync. One button cycles side-by-side, translation-only and original-only; export a translated or alternating bilingual PDF when you're done.
Not a cloud dashboard — a small, native Mac tool that does two things well.
Translation runs on Apple's system engine, right on your Mac. Documents are never uploaded — the first translation of a language pair may download a small pack, managed by macOS itself.
The source language is recognized automatically — or picked manually — and the target defaults to your system language. The supported list comes straight from macOS, with display names in your locale.
Original and translation scroll together in both directions, and zoom mirrors between the panes. One button cycles side-by-side, translation-only and original-only layouts.
Each translated paragraph is drawn back into its original rectangle on a copy of the page. Geometry, margins and vector graphics stay intact — the translation reads like the original.
Math-heavy lines — LaTeX, MathType, Cambria Math fonts — are detected and left untouched, while the prose around them translates. Built for papers and theses: better one line untranslated than an equation destroyed.
Export a translation-only PDF, or an alternating bilingual edition — original page, translated page — for reading anywhere, annotating in Preview, or sharing.
KenjaPDF works like a good tool should — it does its job on your machine and talks to no one. Translation runs on Apple's on-device engine, so your documents never leave your Mac. There is nothing to sign in to, and nothing to leak.
No. Reading, layout analysis and translation all run on your Mac — translation uses Apple's on-device system engine. There are no accounts, no analytics and no tracking; the App Store privacy label is Data Not Collected.
20+ languages, powered by macOS system translation — the supported list comes straight from your system. The source language is auto-detected (or picked manually) and the target defaults to your preferred system language. The first translation of a language pair may download a small language pack, managed by macOS.
No. Translation uses the engine built into macOS — no sign-in, no API keys, no subscriptions. If a language pack is missing, KenjaPDF walks you through the one-time system download and retries.
Yes. Each translated paragraph is drawn back into its original rectangle on a copy of the page, so geometry, margins and vector graphics stay intact. Math-heavy lines are detected and left untouched — formulas survive translation, which matters for papers and theses.
KenjaPDF translates the text layer of a PDF. Image-only scans without embedded text are not translatable in this version — run OCR first if you have a scanned document.
On-device translation requires macOS 26 or later, because it uses Apple's Translation framework. Reading, outline and export run on macOS 13 and later.