How to Translate a PDF Without Losing Formatting

You open a PDF, copy the text into a translator, paste the result into a new document — and an hour later you’re still rebuilding tables. PDF is the hardest common format to translate cleanly, and it helps to understand why before choosing a method.

This guide explains what makes PDFs fragile, compares the realistic options, and gives you a checklist for getting a translated PDF that still looks like the original.

Why PDFs are hard to translate

A PDF is designed for fixed layout, not editing. Internally, a page is closer to a set of drawing instructions than a document with paragraphs:

• Text is stored as runs of glyphs placed at exact coordinates. A “paragraph” might be ten separate lines with no link between them.
• Columns, tables and captions are often just text that happens to be positioned near each other.
• Fonts are frequently embedded as subsets — only the characters used in the original are included. If your translation needs characters that weren’t in the source (Vietnamese diacritics, Polish ogoneks, Cyrillic, CJK), the embedded font simply doesn’t have them.
• Scanned PDFs contain no text at all, only images of pages.

So translating a PDF really involves three jobs: extracting text in the right reading order, translating it, and putting it back into a layout that has to absorb longer or shorter text.

First, check what kind of PDF you have

Try selecting text with your cursor.

• You can select and copy text normally. It’s a “digital” PDF exported from Word, InDesign, LaTeX or similar. This is the good case.
• Nothing selects, or the whole page selects as one block. It’s a scan or an image-only PDF. You need OCR (optical character recognition) before any translation tool can read it.
• Text selects but copies as gibberish. The fonts use custom encodings without proper Unicode mapping. OCR is often the most reliable fix here too.

Also check whether the PDF is password-protected against editing. If it is, you’ll need the owner’s permission or the original file.

Option 1: Translate the source file instead

The most reliable way to translate a PDF is to not translate the PDF. If the document was exported from Word, PowerPoint, InDesign or Google Docs, ask for that original file. Translate it there — the layout engine will reflow text naturally — then export a fresh PDF.

This is always worth one email. It avoids almost every problem below.

Option 2: Convert to an editable format, translate, re-export

If no source file exists, convert the PDF into something editable:

1. Convert to DOCX. Microsoft Word can open PDFs directly and convert them to editable documents; Adobe Acrobat and many other tools can export to Word as well.
2. Inspect the conversion. Look for text boxes where there should be paragraphs, tables that became tab-separated lines, and headers repeated as body text. Fix the structure before translating — it’s much easier in one language than two.
3. Translate the DOCX (our guide on translating Word documents while keeping layout covers this step).
4. Export back to PDF.

This works well for reports, letters and simple brochures. It struggles with heavily designed layouts, where conversions tend to produce dozens of floating text boxes.

Option 3: Use a layout-aware file translator

Tools built for file translation try to do the extraction, translation and re-insertion in one pass: they identify text blocks, translate them in context, and write the translation back into the same positions, adjusting font size or line breaks where the new text is longer.

This is the approach 2Translate is being built around: upload a PDF, choose languages, download a PDF with the same page structure. Whatever tool you use, judge it on the checks in the last section of this guide, not on the translation alone.

Option 4: Rebuild the layout manually

For high-stakes material — a published brochure, packaging, a printed manual — the professional approach is still to translate the text and have a designer typeset it in the original design software. It’s the slowest option and the only one with full control.

Problems you will hit (and how to handle them)

Text expansion

Translations are rarely the same length as the source. Going from English into German, French, Spanish or Vietnamese often makes text noticeably longer; going into Chinese or Japanese usually makes it shorter in characters. In a fixed layout, longer text has to go somewhere:

• Slightly smaller font size in that block (keep it consistent across similar blocks)
• Tighter line spacing
• Shorter phrasing — a good translator can often say the same thing more compactly
• Letting a block grow into empty space below it

Avoid shrinking text below a readable size just to make it fit. If a table cell can’t hold the translation, the table needs to change.

Missing glyphs

If your translated PDF shows empty boxes, question marks or a jarring fallback font, the embedded font doesn’t cover your target language. Pick a replacement font with the needed character set and the closest look to the original. For Vietnamese, for example, check that the font explicitly supports Vietnamese — many Latin fonts lack the stacked diacritics like ệ or ở.

Reading order

Multi-column layouts, sidebars and footnotes can come out of extraction in the wrong order, which produces translations that make no sense because sentences were split across columns. Look at the extracted text before you trust the translation.

Right-to-left languages

Translating into Arabic, Hebrew or Persian mirrors more than text direction: page layout, table column order and alignment usually flip too. Simple text replacement won’t get this right. Budget for manual layout work.

Text inside images

Charts, screenshots, diagrams and logos carry text that no PDF translator will touch, because it’s pixels. Decide which ones matter and recreate them, or add a translated caption.

A quick checklist for the translated PDF

Before you send it anywhere, open the result next to the original and check:

Check	What to look for
Page count	Same as the original, or a deliberate change
Tables	Every cell translated, nothing overflowing or clipped
Fonts	No boxes, no question marks, no mixed fallback fonts
Numbers	Dates, prices and units converted or kept, consistently
Headers and footers	Translated on every page, not just the first
Links	Still clickable and pointing to the right place
Bookmarks and TOC	Translated and still matching the page numbers
Searchability	You can search for a word in the translated text

Which approach should you pick?

• You have the source file: translate that and re-export. Every time.
• Simple text-heavy PDF: convert to DOCX, translate, export.
• Many PDFs or a designed layout you just need readable: a layout-aware file translator, then a careful review.
• Print-quality output: translation plus a designer.

And for anything legal, medical or official, the formatting is only half the job — read our comparison of machine and human translation to decide what level of review the content needs.