How to Translate a PDF Without Wrecking the Layout
Quick answer
Pasting text into a general translator discards the document's structure before translation even starts: columns interleave, tables collapse, and headings lose their hierarchy. Translate PDF works on the file instead — it reads the document, translates the content, and rebuilds it as a PDF with the structure intact. Expect some visual shift, since translated text rarely occupies the same space as the original, and have a fluent speaker review anything that will be signed, filed or published.
Why copy-pasting into a translator falls apart
Selecting the text of a PDF and pasting it into a general translation tool works for a paragraph and fails for a document. Column layouts interleave into nonsense, tables collapse into runs of numbers, headings lose their hierarchy, and what comes back is text with no structure — not a document you can send to anyone.
The problem isn't the translation quality. It's that the document's structure was discarded before the translation ever started, and there's no way to put it back afterwards.
Translating the document as a document
Translate PDF works on the file rather than on pasted text: it reads the document, translates the content into your target language, and rebuilds it as a PDF. You get a document with its structure intact rather than a block of translated prose you'd have to reformat by hand.
Upload the file, choose your target language, and download the rebuilt result. It's a genuine time difference on anything longer than a page — and the difference grows with the complexity of the layout.
Scanned documents need OCR first
A scanned PDF has no text in it, only page images, so there is nothing to translate. Run it through OCR PDF first to add a text layer, then translate the result.
Choose the correct source language when running OCR. It determines the character set and spelling model the recognition uses, and it makes a substantial difference on any language that isn't English — particularly for scripts like Arabic, Devanagari, or Japanese where the wrong setting produces unusable output rather than merely imperfect output.
Text expansion and why layouts shift
Translated text rarely occupies the same space as the original. English to German or Hindi typically expands noticeably; English to Japanese often contracts. That means line breaks land differently and text blocks change height, so some visual shift between the original and the translation is expected rather than a defect.
The practical implication is to check tight layouts — tables with narrow columns, buttons or labels in a diagram, anything with text fitted precisely into a fixed box. Flowing body text handles expansion gracefully; fixed-width elements are where it shows.
Check before you send it out
Machine translation is strong on general prose and weaker exactly where the stakes are highest: legal terms of art, regulatory language, medical terminology, and industry jargon where a term has a specific meaning that differs from its everyday one.
Have a fluent speaker review anything that will be signed, filed, published, or relied on. Have a professional translator review anything legally binding. For internal comprehension — understanding a document someone sent you, or getting the gist of a foreign-language report — machine translation on its own is usually entirely sufficient.
If you need to edit the translation before sending it, PDF to Word converts the result into an editable document so corrections can be made in Word or Google Docs and exported back to PDF.