What it does This repo is a hand-maintained directory of machine-learning libraries, tools, and learning resources that run (or can be ported to) iOS. It catalogs everything from Core ML converters to C++ libraries like dlib and OpenCV, plus Swift-native options such as AIToolbox and MLKit. The author also rounds up web APIs, blog posts, and books for mobile developers who want to do ML without leaving the Apple ecosystem.

The interesting bit The README is refreshingly honest about its limitations: resources are “sorted alphabetically or randomly,” the order “doesn’t reflect my personal preferences,” and some entries are merely “inspiration.” That candor makes it a useful archaeological snapshot of the pre-ONNX, pre-MLX era, when getting a Keras model onto an iPhone still required a Python converter package.

Key highlights

• Covers Core ML converters for Caffe, Keras, scikit-learn, XGBoost, TensorFlow, MXNet, and others
• Tabular comparison of general-purpose libraries with language, license, and dependency manager
• Sections for computer vision, NLP, speech, OCR, game AI, and even “Bioinformatics (kinda)”
• Includes links to pre-trained Core ML model zoos and dynamic model-swapping demos
• Curated learning resources and mobile-focused ML books

Caveats

• Last updated January 12, 2018 — many links and compatibility claims are stale
• Core ML has evolved substantially; the converter landscape looks different now
• Some listed projects are themselves abandoned or have changed names

Verdict Worth a quick browse if you’re maintaining legacy iOS code or writing a history of mobile ML. For greenfield projects, start with Apple’s current documentation instead — this list predates too much of the modern stack to be your primary guide.