Translating a PDF is not as simple as pasting text into Google Translate. The challenge is that PDF is a layout format, not a text format. The content is locked into a fixed visual structure, and translation tools need to extract the text before they can translate it. How well that extraction works determines how clean the translated output is. Here is the full picture of your options.
The Core Challenge with PDF Translation
When you translate a webpage, the browser passes raw HTML text to a translation engine, which returns translated text that slots back into the same HTML structure. PDFs do not work this way. A PDF renderer draws text at specific coordinates on a page. There is no semantic structure like HTML paragraphs or headings, just positioned text strings. Translation tools have to reverse-engineer the reading order, identify sentence boundaries, and reassemble the layout after translating.
For simple single-column text documents, this works reasonably well. For multi-column layouts, documents with text inside tables, or PDFs where the text is embedded in graphics, the results are more variable. Knowing this upfront helps you choose the right method and set realistic expectations about output quality.
Method 1: Google Translate (Upload a PDF Directly)
Google Translate accepts PDF uploads and returns a translated version with layout approximated where possible. This is the fastest free option and works for most standard documents.
- Go to translate.google.com.
- Click the Documents tab at the top of the page.
- Click Browse your computer and select your PDF.
- Select the source language (or leave it on Auto-detect) and the target language.
- Click Translate.
- Google displays a translated version in the right pane. Click Download translation to save the translated document.
The output is a PDF that approximates the original layout with translated text. Images are preserved as-is (including any text that is part of images, which cannot be translated). Google Translate supports over 100 languages for document translation. File size is limited to 10MB for PDF uploads.
Method 2: DeepL (Higher Quality Translation)
DeepL consistently produces more natural-sounding translations than Google Translate for European languages. It also accepts PDF uploads on its free tier, with a file size limit of 5MB.
- Go to deepl.com/translator.
- Click Translate files (the file icon next to the text box).
- Upload your PDF.
- Select the source and target language.
- DeepL translates the document and returns a downloadable translated file.
DeepL's free tier limits the number of documents you can translate per month. The paid DeepL Pro plan removes these limits and adds a glossary feature for maintaining consistent terminology across documents. For technical, legal, or medical documents where precise terminology matters, DeepL Pro is worth considering.
Method 3: Translate via Browser (For Viewing Only)
If you just need to read and understand a PDF in another language rather than produce a translated copy, your browser can translate it on the fly without any upload.
In Google Chrome
- Open the PDF in Chrome by dragging it into a Chrome window.
- Chrome may automatically offer to translate the page if it detects a foreign language. If not, right-click anywhere on the page and select Translate to English (or your preferred language).
- Chrome translates the displayed text. You can read the translated version but cannot download a translated PDF this way.
In Microsoft Edge
Edge has a built-in immersive reader and translation capability. Open the PDF in Edge, click the translate icon in the address bar, and choose the target language. Edge translates the visible content. Again, this is for reading only, not for producing a downloadable translated document.
Method 4: Extract Text First, Then Translate
For longer documents or when layout preservation is not important, extracting the text and translating it separately often produces cleaner results than trying to translate the PDF as a whole. The process is two steps:
- Use our free Extract Text tool to pull all the text from the PDF as a plain .txt file.
- Paste the text into Google Translate or DeepL for translation.
The disadvantage is that you lose all formatting and layout in the output. The advantage is that the translation engine receives clean, unambiguous text rather than text extracted from a positioned PDF structure, which typically produces a more accurate translation. This method works best when you need the translated content as raw text for editing, repurposing, or feeding into another tool.
Method 5: Translate a Scanned PDF
Scanned PDFs are images, which means the text is not machine-readable. Translation tools cannot translate image text directly. You need to run OCR first to extract the text, and then translate it.
- Run the scanned PDF through our PDF OCR tool to create a searchable version with real text.
- Download the searchable PDF.
- Upload it to Google Translate or DeepL for translation.
OCR accuracy depends heavily on scan quality. Poor scans produce OCR errors that carry through into the translation, producing errors that are difficult to identify because they look like valid translated text. For important scanned documents, a human review of the OCR output before translation is good practice.
Which Method Is Right for You?
| Situation | Best Method |
|---|---|
| Simple text PDF, need a translated copy | Google Translate Documents tab |
| European languages, need high translation quality | DeepL file upload |
| Just need to read and understand it | Chrome or Edge browser translation |
| Long document, layout not important | Extract text first, then paste into translator |
| Scanned PDF | OCR first, then upload to Google Translate or DeepL |
| Sensitive document, privacy matters | Extract text locally, translate text only (not the PDF) |
Limitations to Be Aware Of
Machine translation of PDFs has a few consistent limitations worth knowing before you rely on the output:
- Text inside images is not translated. If a chart, diagram, or photo contains text, that text is part of the image and translation tools cannot access it.
- Multi-column layouts often merge incorrectly. Columns may be read left-to-right across both columns rather than reading each column top-to-bottom. The translation itself may be accurate but the reading order can be wrong.
- Tables may lose structure. Table cells can be extracted in the wrong order, especially in complex tables with merged cells.
- Specialized terminology may translate poorly. Legal, medical, and technical documents contain specialized vocabulary that general translation engines sometimes handle inconsistently. Always have an expert review translations of documents with legal or medical significance.
- Google Translate's Documents tab accepts PDF uploads and returns a translated PDF with approximate layout preserved
- DeepL produces better translation quality for European languages; free tier has monthly limits
- Chrome and Edge can translate PDFs on-screen for reading without producing a downloadable translated copy
- Extracting text first with our Extract Text tool often produces cleaner translation input than uploading the PDF directly
- Scanned PDFs need OCR before any translation tool can process them
- Text inside images, multi-column layouts, and complex tables are the weakest points for automatic PDF translation
Need to extract text from a PDF before translating?
PDFToolShack's Extract Text tool pulls all text from any digital PDF instantly. Works in your browser, no account needed.