studio-dots-ai/dots.ocr ? reverse-engineered prompt

Build me a simple web app for OCR and document parsing using this project.

I want to be able to upload an image or PDF of a document, then get back the extracted text in reading order, with the layout kept as close to the original as possible. It should handle multiple languages, tables, charts, diagrams, and messy scanned pages without me having to tweak anything. If the model can also show bounding boxes or a visual preview of detected blocks, that would be great.

Please make the interface clean and easy to use, with a drag and drop upload area, a results panel, and a way to download the parsed output as text or structured data. If there’s already a demo flow in the repo, use that and make it polished. Look up current docs online if you need to.