← All guides

Translate your first PDF

A PDF is the file people are most reluctant to paste into a website — a contract, an invoice, a medical result. This is how you translate one without it ever leaving your Mac.

  1. 1Add a PDF and check the engine

    Documents require Pro or an active trial. Open the files panel and add a PDF. For this guide, choose an on-device translation engine and install its language resources before starting.

    The document panel with a PDF loaded
    Historical interface illustration; capture version unverified. It does not demonstrate current projects or glossary correction.

  2. 2Pick the language pair

    Choose the source and target language above the panels. Leave the source on automatic if you are not sure — the app detects it from the first page of text.

    The language pair selected above the two panels
    Historical interface illustration; capture version unverified. It does not demonstrate current projects or glossary correction.

  3. 3Translate and review progress

    Start translation and watch the reported progress. In a normal project, saved work restores paused. Resume explicitly; completed stages are retained and the smallest interrupted stage may repeat. Private sessions do not persist content. PDF output is text, not a reconstructed PDF.

    Translation running with per-section progress and an ETA
    Historical interface illustration; capture version unverified. It does not demonstrate current projects or glossary correction.

  4. 4Check a page you care about

    Review the translated text alongside the source PDF, including OCR accuracy and terminology. Glossary handling depends on the selected engine; check the result after any changes.

    The translated result next to the original
    Historical interface illustration; capture version unverified. It does not demonstrate current projects or glossary correction.
Recording not captured
The whole run, from dropping the file in to the finished translation.

What about scanned PDFs?

Pages without a text layer go through on-device OCR automatically — Apple’s Vision framework first, with a bundled Tesseract fallback for the languages Vision does not read, Polish among them. Mixed documents work too: only the pages that need OCR get it. No OCR model is downloaded and nothing about OCR touches the network.

That is the whole loop

DOCX, PPTX and DRM-free EPUB can also be translated. Their output formats differ from PDF text extraction. Keep the on-device engine selected to process translation on this Mac.

Coming soon to the Mac App Store