Translating a Hindi PDF to English Without Retyping: What Actually Works

Javed Bhati — Tue, 27 Oct 2026 22:29:56 +0000

Anyone who has tried to copy text out of a Hindi PDF knows the frustration: you paste it somewhere and the letters fall apart, matras land in the wrong place, or half the document turns into boxes. Retyping a 20-page marksheet or certificate is nobody's idea of a good afternoon.

The reason copy-paste fails is usually the font. Many Hindi PDFs use legacy fonts (like Kruti Dev) instead of Unicode. The text looks perfect on screen, but underneath, the characters aren't stored as proper Hindi — so nothing you paste into Google Translate comes out right.

Here is the approach that actually works:

Check whether your PDF has real text. Try selecting a line. If it selects, it is a text PDF; if nothing selects, it is a scanned image and needs OCR first.
For scanned documents, run OCR before translating. An OCR step that understands Devanagari converts the image into real, selectable Hindi text.
Translate the document as a whole, not line by line. Translating in chunks breaks tables, headers and signatures out of order. A document translator keeps the layout intact — you get the same pages back, in English.
Always proofread names and numbers. Machine translation is good at sentences but can fumble proper nouns, roll numbers and dates. A two-minute check of those fields saves an embarrassing resubmission.

I wrote a longer step-by-step walkthrough of this process here: how to translate a Hindi PDF to English for free

The short version: do not fight the font, do not retype, and do not translate line by line. Get the text right first, then translate the whole document.

DEV Community: Javed Bhati

Translating a Hindi PDF to English Without Retyping: What Actually Works