AI Translator

AI Translator extracts text from your PDF using pdf.js, then uses a local translation engine powered by WebAssembly to translate the content. The translated text is placed back into the PDF structure, preserving the original layout, fonts, and formatting. Everything happens in your browser — no data is sent to any server.

Over 100 languages are supported, including all major European languages, Chinese (Simplified and Traditional), Japanese, Korean, Arabic, Hebrew, Hindi, Bengali, Thai, Vietnamese, Turkish, Russian, and many more. Right-to-left languages like Arabic and Hebrew are fully supported with correct text direction.

Yes. The translator preserves the original layout — paragraphs, columns, tables, images, and page structure all remain in place. Font matching automatically selects fonts that support the target language's character set, so CJK, Arabic, and Cyrillic text all render correctly.

Yes. Scanned PDFs are automatically processed with OCR before translation. Text is extracted from the scanned images and then translated — all in one pass. You don't need to run a separate OCR step first.

No. AI Translator runs entirely in your browser using JavaScript and WebAssembly. Your documents — contracts, research papers, medical records — are processed on your device and never transmitted anywhere. This is especially important for confidential or legally sensitive documents.

Translation accuracy depends on the complexity of the source text. General business and personal documents achieve high accuracy. Technical, legal, or highly idiomatic text may require human review — the quality score feature helps you identify pages that benefit from additional attention. Using a custom glossary improves consistency for domain-specific terminology.

Yes. Switch to batch mode to queue multiple PDFs and translate them all to the same target language. This is perfect for translating a series of related documents. Download all translated files individually or as a ZIP archive.