DeepSeek OCR — распознавание текста в документах

★ 1.9K

deepseek_ocr_app is an open-source ready application for text recognition built on the DeepSeek-OCR model. Reach for it when you want a working recognition tool with an interface rather than a library to embed: upload an image or a multi-page PDF up to 100 MB and get the recognised text. Built as a web app with a React front end and a FastAPI back end, it deploys as your own service and a whole team uses it through the browser. The difference from raw OCR models is that this is a finished application over a strong modern model: no wrapper code to write, it is all assembled. It suits digitising documents inside a company without sending files to external services. Recognition accuracy on difficult scans is worth checking on your own documents.