Media
4 items
Description
Image-based translation with spatial text placement and inferred reading order. AI Context-Aware Image Translator Read text inside images in another language. Translate comics, picture books, scanned pages and image-based documents through an AI connection you choose, with translations displayed near the original text. TRANSLATE A PAGE OR AN AREA Translate page images, or select an area you want to read. Open supported images and PDFs in the extension's viewer. Recognition and placement depend on the source image, page layout and model you configure; some websites or protected viewers may prevent image access or capture. KEEP THE CONTEXT Use recent reading context and a per-work Series DB to guide names, terms and expressions. Manage the work's reference information yourself and export or import it as a file. Keep your reading choices available across pages without a developer account. CHECK THE ORIGINAL AND MAKE CORRECTIONS Inspect recognized source text, compare the original with the translation, and ask your configured AI for a short explanation. In selected-area reading, save source corrections, translation edits and font adjustments locally for matching source content. BRING YOUR OWN AI The extension is free. Configure an image-capable local model or supported API endpoint for recognition, and a text-capable connection for translation. You can use the same suitable connection for both. API and model-hosting charges, if any, are separate. This image extension does not use the optional subscription helper offered with our text-based products. KNOW WHERE YOUR DATA GOES Image pages or selected crops go to the recognition connection you configure. Recognized text and relevant reading context go to your configured translation connection. Settings, API credentials, reference data and saved edits are stored locally in the browser. Optional chapter preparation may read supported next pages from the source website. There is no developer-operated translation server, developer sign-in, analytics or telemetry. A fully local model can process AI requests on your computer; a remote API processes content under that provider's terms and privacy policy. Website requests remain separate from model processing. The interface supports English, Korean, Japanese, Simplified Chinese, Spanish, Brazilian Portuguese, French and German. Translation quality and available target languages depend on your chosen model.