Translate PDF

Translate a PDF

Translate your PDF into another language and keep the original layout intact.

Translate a PDF online

PDF Vision translates a PDF and rebuilds the page instead of returning a wall of text. Each page is first rendered and placed as a full-page background, so images, logos, rules and colours are preserved exactly where they were. Every text block is then read with its position and font size, translated into the language you pick, and redrawn as real, selectable text inside the box the original occupied, shrunk to fit when the translation runs longer. Unicode fonts are embedded so Chinese, Japanese, Korean, Cyrillic, Hindi, Thai and Greek render as real glyphs. Because that background is a rendered image, non-text elements are no longer vector and can look softer when you zoom or print. You get the translated PDF plus the original and translated text side by side.

How to translate a PDF

  1. Upload your PDF

    Sign in with a Premium or Gold plan and upload a PDF of up to 10 MB that contains real, selectable text.

  2. Pick the languages

    Leave the source on auto-detect or set it manually, then choose one of the nineteen target languages.

  3. Download the translated PDF

    Check the side-by-side comparison of original and translation, then download the rebuilt PDF or copy either column.

Why use it

Layout survives

Translated sentences are redrawn at the coordinates of the originals, with the font size reduced when the new wording runs longer, so tables, columns, headers and page breaks keep their shape. That shrinking goes down to a 4 pt floor, so a very expansive language pair can end up noticeably smaller than the original.

Non-Latin scripts render

A Unicode font is embedded for the script you translate into, so Chinese, Japanese, Korean, Cyrillic, Devanagari, Thai and Greek appear as real characters rather than blank boxes or question marks.

PDF and plain text

Alongside the rebuilt document you get the extracted source text and the translation as plain text, each copyable in one click for pasting into an email, a CMS or a translation memory.

Frequently asked questions

Is translating a PDF free?
No — translation needs a Premium or Gold plan, because every page is sent to an AI provider and billed. There is no extra per-page or per-word charge on top of the plan: Premium includes 20 AI documents a day, Gold 200. Translation keeps its own 10 MB ceiling per PDF, below your plan's general limit, because a translated page costs far more to process than a compressed one.
Which languages can I translate into?
Nineteen targets: English, French, Spanish, German, Italian, Portuguese, Dutch, Swedish, Polish, Greek, Russian, Turkish, Arabic, Hindi, Chinese, Japanese, Korean, Vietnamese and Thai. The source language can be auto-detected or set by hand, and once it is set by hand the two sides can be swapped in one click.
Can I translate a scanned PDF?
Not directly. The tool reads the text layer of the PDF, and a scan is only a picture of a page. Run OCR on it first to create that text layer, then translate the searchable file it produces.
Does Arabic come out correctly?
Arabic is available and translated accurately, but the rebuilt PDF places right-to-left text without full letter joining and reordering, so the page can look wrong. Copy the translation from the text panel instead, which is correct.
Will the translated PDF look identical to the original?
Very close. Each page's original appearance — images, logos, rules, colours — is preserved as a rendered background, and the translation is drawn on top as real, selectable text inside the box the original text occupied. Because that background is rendered rather than kept as vector, non-text elements can look softer when you zoom or print, and the file is often heavier than the source. Anything that could not be reproduced is listed in a note above the result.
Does the file stay in my browser?
No. The PDF is uploaded to our servers so the text can be extracted and translated. The tools that never send your file anywhere are the browser-only ones: merge, split, rotate, watermark, edit and JPG to PDF.

Related tools