Read difficult page layouts
Detect text across multiple panels, including compact bubbles and vertical Japanese dialogue.
Text recognition · Translation · Typesetting
Recognize dialogue from manga images and turn it into a finished translated page. Review the result visually instead of managing a separate list of extracted text.
Quick answer
A manga OCR translator combines optical character recognition with translation and page lettering. It reads text inside speech bubbles or captions, translates the recognized lines and renders editable text in the corresponding area.
Designed for manga pages
Detect text across multiple panels, including compact bubbles and vertical Japanese dialogue.
Skip manual transcription: recognized lines move directly into the translation and typesetting workflow.
Use the visual editor to correct text, resize the box or reposition a line before export.
How it works
Drop a manga scan or screenshot into the browser.
Choose the source and target languages, then process the image.
Open the completed page, edit any line and export the image.
FAQ
Yes. The recognition workflow supports vertical Japanese dialogue as well as horizontal text.
Yes. Recognition is used behind the page workflow, and the finished translated image is the main result.
The main source languages are Japanese, Korean, Chinese and English, with automatic per-page language detection.
Yes. Open the image editor to correct translated text and adjust its size, width or position.
Choose your workflow
Start with an image task or open a translator already set to your language pair.