Samedoc / Google Translate PDF not working
Google Translate won't translate your PDF? Here's why.
Five things stop Google Translate's Documents tab, and each one has a fix. If your PDF is a scan, a payslip or a bank statement, the quickest fix is a translator that reads the page itself.
Updated November 6, 2026
Why it fails, and the fix
Google Translate translates PDFs up to 10 MB and 300 pages, on a computer, and only the text it can read. It skips anything that is a picture: scans, phone photos, and PDFs whose text is stored in a broken way. That's the most common reason for "Can't translate scanned PDFs", or for a download that comes back still in the original language.
| What you see | Why | Fix |
|---|---|---|
| "Can't translate scanned PDFs" | The PDF is a scan or a photo. The words are pixels, not text. | Use a translator that reads the page image (Samedoc does). OCR first also works, but you lose the layout. |
| The same message on a normal PDF, or a download still in the original language | The PDF's text is stored with a broken character map. Common in payroll and bank PDFs from older software, especially Hebrew and Arabic. | Translate from the page image, not the text layer. |
| The upload is rejected | The file is over 10 MB, or the PDF has more than 300 pages. | Compress or split the file, or translate only the pages you need. |
| No Documents tab | Document translation isn't available on small screens or phones. | Use a computer, or a translator that takes phone photos. |
| The file type isn't accepted | Google takes only .docx, .pdf, .pptx and .xlsx. | Save or export the file as PDF or DOCX first. |
Limits from Google Translate Help, checked November 6, 2026.
What we saw on a real payslip
In November 2026 we ran a real Israeli payslip through Google Translate, DeepL, iLovePDF, onlinedoctranslator and Samedoc. The file was a normal digital PDF, yet Google refused it as a scan. The payroll software had written most labels in fonts with a broken character map: only 678 of its 1,710 characters came out as Hebrew, the rest as symbols. Tools that read the text layer see symbols, not words.
Samedoc reads the page the way you see it, so broken text doesn't matter. Below: the original and Samedoc's English copy, unedited.
Original (Hebrew)
Samedoc (English)Translate the PDF Google won't
- Upload the originalA PDF, a scan or a phone photo (PDF, PNG, JPG or WEBP, up to 25 MB).
- Pick the languageHebrew, Arabic and other right-to-left pages included. You can switch the layout to left-to-right.
- Download the same documentA PDF in the original layout, with a quality report. Need to edit it? Turn it into Word with PDF to Word.
Questions
Why does Google Translate say "Can't translate scanned PDFs" when my PDF isn't scanned?
Google couldn't read the text inside your PDF. That happens with scans, and also with digital PDFs whose fonts store letters in a broken way, which is common in payslips and bank statements in Hebrew and Arabic. A translator that reads the page image works on both.
What is Google Translate's PDF size limit?
10 MB per file, and PDFs of up to 300 pages. Google's supported file types are .docx, .pdf, .pptx and .xlsx.
Can I translate a PDF with Google Translate on my phone?
Not as a document. Google says document translation isn't available on smaller screens or on mobile. The camera mode shows a translation on screen, but it doesn't give you a translated file.
Is it free?
The free plan includes 1 document a day, up to 3 pages. For longer documents or a packet of up to 200, see bulk document translation.
What happens to my file?
It is used only to make your translation and is deleted after 24 hours. See the privacy policy.
Translate the PDF Google won't
Scans, payslips and bank statements. 1 free document a day, same layout.