A new post on the Hugging Face blog introduces dots.ocr as a state-of-the-art optical character recognition (OCR) system built for Apple's Core ML framework. The title frames the work as a way to run high-quality text recognition on Apple hardware, with Core ML serving as the underlying machine learning runtime.
Because this article is based on a single source, there are no independent accounts to weigh against each other. The "state-of-the-art" label comes from the post's own framing, and the available information does not include benchmark numbers or architectural details.
Developers evaluating OCR options for Apple platforms may want to read the original post for code, model details, and any performance comparisons it provides.