OCR in Many Languages – Even Mixed on One Page
The OCR engine behind OCR PDF, Tesseract, recognizes each language with its own trained model. After scan quality, choosing the right languages is the biggest factor for accuracy: accented letters, umlauts, Cyrillic, Greek, Arabic and Asian scripts are only read correctly when their language is selected. You can choose up to four languages for one document.
How to Recognize a Multilingual Document
- Upload the file Open OCR PDF and choose a scanned PDF or photos of the pages.
- Add every language Click "+ Add language" for each language in the document; remove the ones that are not used.
- Start OCR Pages with several languages are recognized in one pass.
- Check the special letters Look at names and words with accents or other scripts in the recognized text.
Languages and Scripts
Available languages
العربية, বাংলা, Български, Čeština, 简体中文, 繁體中文, Deutsch, Ελληνικά, English, فارسی, Français, עברית, हिन्दी, Magyar, Bahasa Indonesia, Italiano, 日本語, 한국어, Bahasa Melayu, Nederlands, Polski, Português, Română, Русский, Español, Svenska, ไทย, Türkçe, Українська, اردو, Tiếng Việt.
Accents and special letters
Letters such as é, ß, ñ, ø or ł are only recognized reliably when their language is chosen. With English alone they often turn into similar plain letters.
Bilingual documents
Contracts with an annex in another language, multilingual forms and dictionaries: choose both languages and they are recognized on the same page.
Right-to-left scripts
Arabic and other right-to-left text keeps its reading order in the copied text.
Chinese, Japanese and Korean
Choose the exact script: Simplified and Traditional Chinese are separate models.
Fewer is better
Each extra language makes OCR slower and can confuse similar letters, so choose only the languages that really appear.
Good to Know
- Old typefaces such as German Fraktur and historical spellings are recognized poorly by modern language models.
- Handwriting is recognized much less reliably than printed text in every language.
Multilingual OCR – Frequently Asked Questions
How many languages can I use at once?
Up to four per document. They can appear on the same page.
Which languages are available?
العربية, বাংলা, Български, Čeština, 简体中文, 繁體中文, Deutsch, Ελληνικά, English, فارسی, Français, עברית, हिन्दी, Magyar, Bahasa Indonesia, Italiano, 日本語, 한국어, Bahasa Melayu, Nederlands, Polski, Português, Română, Русский, Español, Svenska, ไทย, Türkçe, Українська, اردو, Tiếng Việt.
My language is missing – what can I do?
The list shows every language installed on our OCR server. Contact us and tell us which language you need.
Why are accents missing in my text?
The language of the text was probably not selected. Run OCR again with that language added.
Is it free?
Yes, without sign-up or watermark. Files are deleted after processing, after one hour at the latest.
Recognize Text in Your Language
Choose up to four languages and make your scanned PDF searchable with OCR PDF.
Start Multilingual OCR