OCR

Recognize text in images across 23 languages — the same engine that powers our translators.

Drop an image here, or click to choose
Extract text without translating · JPG/PNG/WebP · ≤10MB

Heavy pages (manga) can take around a minute — speedup in progress.

See it in action

Beforebefore — OCR example
Afterafter — OCR example

Example assets are placeholders — being replaced before launch.

How to use

Three steps, one clean result

1

Upload your image

Drag & drop, browse or paste (Ctrl+V) any image with text in it.

2

Pick a language

Choose from 23 languages — the source language is detected automatically.

3

Download the result

Get a PNG with text translated in place — same layout, same style.

Why OCR

23 languages

CJK, Cyrillic and European scripts recognized out of the box.

Layout aware

Detects text regions as clusters — paragraphs, bubbles, captions — not a wall of characters.

Built on the same engine

The detection layer our manga translator uses, exposed as its own tool.

2 free a day

Every account extracts 2 images a day free — paid plans include unlimited OCR.

FAQ

Questions, answered

How long does a translation take?+

Most images finish in 15–60 seconds. Dense manga pages can take a little longer — we're shipping speedups continuously.

Which languages are supported?+

23 languages: English, Chinese, Japanese, Korean, Russian, 14 European languages (Spanish, French, German, Portuguese, Italian, Dutch, Polish, Czech, Swedish and more), plus Vietnamese, Thai and Arabic.

What file formats and sizes are supported?+

JPG, PNG and WebP, up to 10MB per image. You can drag & drop, browse, or paste (Ctrl+V) directly.

Ready to translate your first page?

2 free images a day — no watermark.

Start Translating