Translate PDF files online

Try selecting a sentence in your PDF reader. A document exported from Word often has a text layer, while a scan is usually an image of a page. Some scans contain OCR text, so this is a useful clue rather than a perfect test. The upload process determines how the document can be read.

For a supported text-based PDF, the file is converted to an editable document, translated and rebuilt. The goal is a layout-faithful result rather than a pixel-identical copy. Paragraphs and table cells can reflow because translated text has a different length.

Scanned pages are read before translation. The recognised text is used to create a clean new document, retaining supported headings, lists and tables where possible. This is a reconstruction, not an exact reproduction of the original page design.

Use straight, well-lit pages with sharp text and complete page edges. Check names, reference numbers, dates and any uncertain reading carefully. Handwriting, stamps and poor images can be difficult to interpret accurately, even when the document appears readable to a person.

• If you have the original DOCX, XLSX or PPTX, upload that instead when preserving editable structure is the priority.
• PDF uploads are limited to 20 MB and 10,000 words. Scanned or photographed jobs support up to 20 pages.
• Review the quoted word count and any reading warnings before confirming a scanned job.
• Download both the PDF and editable DOCX. Use the DOCX if you need to adjust text, tables or page breaks before exporting again.
• Keep a local copy within the 30-day file access period. AI translation does not include certification or human review.