How to translate a PDF into another language
Updated
To translate a PDF into another language, open the Translate PDF tool on pdf.xyz, pick your file, and choose the language you want. AI translates the PDF's selectable text and returns it as translated Markdown and plain text, ready to download.
Translate PDF is one of the AI tools, so the work does not happen in your browser: your PDF uploads over an encrypted connection to an encrypted, isolated worker, and inputs are wiped when the job finishes. The download link stops working after 60 minutes, and stored copies are erased by a daily storage rule. No account is needed at any point.
Know what you are getting before you start: a text translation, not a page-by-page rebuild. The output is the document's words in the new language, packaged as Markdown and plain text rather than a re-laid-out PDF. That focus on the text is also what lets long documents be handled sensibly — split at page boundaries, translated in ordered passes, and reassembled rather than refused.
Open Translate PDFStep by step
- 1Open the Translate PDF tool.
- 2Pick the PDF you want translated. The tool reads selectable text, so if your document is a scan, run it through OCR PDF first.
- 3Upload your PDF — it travels over an encrypted connection to the secure worker that runs the job.
- 4Choose your options, including the language you want the text translated into.
- 5Start the translation. A long document is not refused: it is split at page boundaries, translated a few pages at a time, and reassembled in order. The worker deletes the input when the job finishes.
- 6Download the result — translated Markdown and plain text, packaged together. The download link stops working after 60 minutes, so save it to your device.
What you get back
Translate PDF outputs a Markdown file and a plain-text file, delivered together as a single ZIP. Markdown keeps the translation usable as a document — read it as-is, convert it, or paste it wherever the text needs to go. Plain text is the same words with nothing else attached.
What it does not produce is a re-laid-out PDF. The translation covers the words, not the design: fonts, columns, and positioning from the original are not reconstructed in the new language. If the goal is the same PDF in a different language, plan on taking the translated text into your own layout — this tool's job is getting the words right. If what you need from a long document is the gist rather than a full translation, Summarize PDF works from the same selectable text.
Long documents, oversized pages, and the allowance
Each translation pass carries a bounded amount of selectable text, because translation writes roughly as much as it reads. The exact character figures, and the approximate page ranges they work out to, are stated on the Translate PDF page itself. A longer document is not refused: it is split at page boundaries, translated a few pages at a time, and reassembled in order, so the result reads front to back like the original.
Two limits can still show up in the output, and both announce themselves. A single page holding more text than one pass can carry cannot be translated at all — it is skipped, and the result says so in place, right where that page was. And the amount translated in a day is bounded by the free allowance. 30,774 free characters of AI text a day for your connection — the allowance resets daily. A document longer than the allowance covers is translated as far as the allowance reaches, with the rest marked in place and the reason named — nothing stops silently. Files themselves are capped at 100 MB.
Where your file goes
Translate PDF is one of the AI tools, so processing does not run in your browser. Your PDF uploads over an encrypted connection to an encrypted, isolated worker, which is where the AI reads the text and writes the translation. Inputs are wiped when the job finishes. The download link stops working after 60 minutes, and stored copies are erased by a daily storage rule.
No account is needed for any of this. Open the tool, translate, and download the result.
Scanned PDFs need OCR first
The tool translates a PDF's selectable text — the kind you can highlight and copy. A scanned document is a picture of text, so there is nothing selectable to translate. Run the scan through OCR PDF first to recognize the text, then bring the result to Translate PDF.
OCR PDF advertises 21 recognition languages, with up to 4 selectable per job, and its page's live status shows which language packs are actually installed.
Common questions
Can I translate a PDF for free?
Yes, with no account. 30,774 free characters of AI text a day for your connection — the allowance resets daily. If a document needs more than the allowance has left, it is translated as far as the allowance reaches, and the untranslated remainder is marked in place with the reason named.
What happens to my PDF after it is translated?
The file is processed on an encrypted, isolated worker, and inputs are wiped when the job finishes. The download link stops working after 60 minutes, and stored copies are erased by a daily storage rule.
Will the translation look like the original PDF?
No. The output is translated Markdown and plain text, not a re-laid-out PDF. The words are translated; the original's design is not rebuilt around them. Take the text into your own document if you need the translation typeset.
How much of a PDF can be translated in one run?
Each pass carries a bounded amount of selectable text — the exact figures are stated on the Translate PDF page. A longer document is split at page boundaries, translated pass by pass, and reassembled in order. Two things bound how far one run gets. A single page holding more text than a pass can carry cannot be translated at all — it is skipped, and the result notes it in place. And the free daily allowance: the run translates as far as the allowance reaches, with the rest marked in place and the reason named.
Can I translate a scanned PDF?
Not directly — a scan has no selectable text. Use OCR PDF to recognize the text first, then translate the output.