PDF Translator

Translating a PDF is fundamentally harder than translating plain text. The text is positioned at specific coordinates on each page; the translated text has a different length than the source, which means the layout has to reflow. Translating 'Good morning' (12 characters) to 'Buenos días' (11 characters) is roughly fine. Translating 'Don't' (5 characters) to 'Ne pas' (6 characters) is fine. But translating a German compound word to its English equivalent can shrink or grow by 3-5×, which breaks tight layouts.

A good PDF translator preserves the visual structure (page size, image positions, headings, lists) while allowing the text to reflow naturally within each text block. Some reflow is unavoidable — that's the cost of translation. But the page structure should remain recognizable: a 5-page contract stays a 5-page contract (or maybe 6), a slide deck keeps one slide per page, a brochure keeps its multi-column layout.

For legal, medical and high-stakes content, machine translation is a starting point, not a finished product. Use it to understand the gist of a foreign-language document; for binding use, get a human translator. We'll ship this tool once the layout-preservation pipeline is reliable across a range of document types and language pairs.