EDGE FDE
BeginnerCore AI vs Core ML vs MLX — Which to Use
Learn which Apple stack to use for an on-device model. Core ML, Core AI, and MLX are not the same tool.
For someone who is new to Apple on-device AI and wants a clear pick before they train or ship.
Lessons
Key concepts
- Core ML, Core AI, and MLX are three Apple tools. They are not the same stack.
- Core ML ships classic models in apps. Core AI ships custom neural or generative models on OS 27 or later.
- MLX trains and tests on a Mac. You export before you ship.
- Foundation Models is Apple's system language-model API. It is not a path for your own trained file.
Takeaways
- You can pick one stack from OS target, model job, and whether you are training or shipping.
- You can say one job for each name without swapping them.
- You know the sibling courses for shapes, catalog presets, and the PyTorch to Core AI build.
Agent briefs
FAQ
What is the difference between Core AI, Core ML, and MLX?
They are three Apple tools, not one stack. Core ML ships classic ML models in apps. Core AI ships custom neural or generative models on iOS/macOS 27+. MLX trains and tests on a Mac; you export before you ship.
When should I use Core ML?
For classic ML (classifiers, trees, regression) and apps that must run below iOS 27 / macOS 27. Convert with Core ML Tools to .mlmodel and load via Core ML.
When should I use Core AI?
To run your own neural / generative model on-device with a .aimodel on iOS 27+ / macOS 27+ and Xcode 27. Train in PyTorch, convert with coreai-torch (see PyTorch to Core AI in Xcode).
When should I use MLX?
To train, fine-tune, or experiment on Apple silicon Macs. It is not the main App Store inference API — export to Core ML or Core AI before shipping.
Is Foundation Models the same as Core AI?
No. Foundation Models is Apple’s system language-model API. Core AI is the path for your own trained / converted model file.
Related courses
PyTorch to Core AI in Xcode · Core AI models, typed · Route work to the right Apple model · MLX Swift in production · Run One Large LLM Across Several Macs with MLX