We dropped a scanned PDF into Google Translate's document tab and it refused outright. Here's why — and the workflow that actually translates a scan.
If you've tried to translate a scanned document with Google Translate and hit a wall, you're not imagining it. We uploaded a scanned page from the US Dietary Guidelines to Google Translate's document tab — the "Documents" upload on translate.google.com — and it refused outright. No translated file, no OCR, nothing to download. Just a small message in the corner:
Notice the wording — "Can't translate scanned PDFs." Google itself recognizes the file is a scan and stops there, rather than even trying. The short answer to "can Google Translate a scanned PDF?" is therefore: not directly. To understand why — and what to do instead — it helps to know the one thing Google Translate doesn't do.
Google Translate's document feature only reads the text layer that's already inside a PDF. A normal, digitally-created PDF (exported from Word, generated by software) has that layer — the actual characters underneath what you see. Google grabs those characters, translates them, and hands back a translated file.
A scanned PDF has no text layer. It's a photograph of a page wrapped in a PDF — to software, every word is just pixels in an image. Turning that picture back into real characters requires OCR (optical character recognition), and Google Translate's document tab doesn't run OCR on your upload. With no text layer to read and no OCR to create one, it has nothing to translate, so it stops.
Open your PDF and try to select a sentence with your cursor. If you can highlight individual words, it has a text layer and Google Translate's document tab will probably handle it. If your selection grabs the whole page like a picture (or nothing), it's a scan — and that's why Google refuses it.
It's worth being precise here, because "Google can't translate PDFs" isn't quite true:
So if you have a scanned contract, certificate, or transcript as a PDF, the Documents tab is a dead end — and the camera trick won't give you a tidy, layout-correct document back.
Google Translate can get you a translation of a scan, but only if you add the OCR step yourself and tidy up the layout afterwards. Here are the two routes side by side:
| Google Translate | Reglyph | |
|---|---|---|
| Upload the scan directly | Refused with "Can't translate scanned PDFs" | Yes, including PDFs from a phone scanner app |
| No separate OCR step | Run an OCR tool first (e.g. Acrobat's "Recognize Text") to make a PDF or Word file with text | Recognized automatically |
| No layout fixing afterwards | Tables, stamps and signature blocks often shift or go missing; fix them by hand against the original | Translation goes back where the text was |
| Steps | 3: OCR → upload and translate → fix the layout | 1: upload |
| Tools needed | An OCR tool, Google Translate, often Word too | Just Reglyph |
For a single short page where you only need the gist, the Google route works. For anything long or structured, which covers most official paperwork, step 3 is where the time goes and where mistakes creep in.
Reglyph folds the three steps into one: it reads the text on the scan, translates it and sets it back where it was, so tables, columns, figures, stamps and numbers stay in place. 5 pages a day are free: try it here.
Can Google Translate a scanned PDF? Not directly — its document tab only reads PDFs that already contain text, and it won't OCR a scan, so image-only documents get refused. You can OCR the file yourself first and then translate it, but you'll lose the formatting and have to rebuild it. If you want to skip all of that, a translator built for scanned PDFs does the OCR, translation, and re-typesetting in one pass and gives you back a clean, layout-correct document.
For how Google Translate stacks up against DeepL, ChatGPT and five other tools on the same scans and papers, see our PDF translation tools comparison.
Upload a scanned PDF — even one made with your phone's scanner app. Reglyph OCRs it, translates it, and rebuilds the page so tables, stamps, and figures stay exactly where they were.
Translate 5 pages free, every dayNot directly. Google Translate's document tab only reads PDFs that already have a text layer, and it doesn't run OCR on uploads. A scanned PDF is an image with no text layer, so Google Translate refuses it or returns nothing. You need OCR first — or a tool that OCRs the scan automatically.
Because your PDF is most likely a scan — an image of a page with no underlying text. Google Translate has no characters to read and won't convert the image to text itself, so it stops with an unsupported or 'can't translate' message instead of producing a file.
Run the scan through an OCR tool first (Acrobat's 'Recognize Text' or a free OCR site) to create a PDF or Word file that contains real text, then upload that to Google Translate. The downside is you lose the original layout and have to rebuild tables and columns by hand.
Yes — use a translator built for scans that OCRs the page, translates it, and re-typesets the result back into the original positions. Reglyph does this in one step: upload the scanned PDF as-is and download a translated PDF with tables, figures, and stamps still in place.
Vision models genuinely read scans now. The question isn't whether ChatGPT can translate your PDF — it's whether a chat reply is actually what you need back.
Read guideCamera-translate apps are great for reading a page on the spot, but they don't give you a file — here's how to get an actual translated PDF from your phone.
Read guideRandom symbols, blank pages, or 'PDF tools are temporarily offline' — the cause is almost always the same one thing, and it's fixable.
Read guideSimple, scan-friendly pricing. Pages are pages — no multiplier for scanned files.
Translate a few pages. No commitment.
$0.30 a page, from 10 pages — you pay for exactly 10 pages.
For individuals with scans to translate every month.
For high-volume users with scans to translate every week.