meangrinch/MangaTranslator ? reverse-engineered prompt

Build me a web app that can take manga or comic page images and automatically translate the text inside the speech bubbles and any text outside the bubbles too.

I want it to feel like a one click tool, where I can upload a single image or a whole folder or zip file, then get back cleaned up translated pages with the text redrawn in the right places. It should detect the text areas, remove the original text, translate it into lots of languages, and render the new text with good looking fonts. If possible, make it support custom font packs so I can drop in my own fonts for different languages.

Please include a simple web UI and a command line option too. It should work with GPU if available, but still have a CPU mode. Also make it save outputs neatly and keep the original folder structure when processing batches.

If you need to check current docs for any AI or model setup details, go ahead and look them up online.