---
title: "Pick a stack with a table"
lesson_id: "05"
---

# Pick a stack with a table

- **Lesson ID:** 05
- **Goal:** Give a decision table, a checklist, and links so the reader can pick Core ML, Core AI, or MLX from OS, model, and job.
- **Human lesson:** [05-choose-with-confidence.html](05-choose-with-confidence.html)

## Prerequisites

- Lessons 01 to 04, or a working idea of each tool's job.

## Inputs, outputs, and artifacts

- **Inputs:** Lowest OS version, model job, and whether the reader is training or shipping.
- **Outputs:** One stack pick and a next course.
- **Artifacts:** Progressive `.flow` diagrams, a decision table, a checklist, and source links.

## Agent build steps

1. Start from OS, model, and output, not from a brand name.
2. Put four rows in the table: Core ML, Core AI, MLX then export, and Foundation Models.
3. Repeat the short form: older OS or classic ML means Core ML. Custom neural or generative work on 27 or later means Core AI. Train on a Mac with MLX, then export to ship.
4. Give a numbered checklist the reader can follow before they start work.
5. Link [Model architectures in plain English](../model-architectures-plain/), [Core AI models, typed](../coreai-models-typed/), and [PyTorch to Core AI in Xcode](../pytorch-core-ai-xcode/).
6. Cite [developer.apple.com/core-ai](https://developer.apple.com/core-ai/), [apple/coreai-models](https://github.com/apple/coreai-models), and [MLX](https://github.com/ml-explore/mlx).

## Constraints

Keep the table factual. Do not invent deprecation. No quizzes, Mermaid, or em dashes.

## Key concepts

- Start from OS, model job, and output, not a brand name.
- Older OS or classic ML: Core ML, `.mlmodel`, `coremltools`.
- Custom neural or generative on 27 or later: Core AI, `.aimodel`, Xcode 27.
- Train or shrink on a Mac: MLX, then export before you ship.
- Apple's system LLM: Foundation Models, not your trained file.

## Takeaways

- You can run the checklist and get one stack pick before you start work.
- You can justify Core ML versus Core AI versus MLX versus Foundation Models.
- You can check that file type and OS match before you build.

## Acceptance checks

- The table covers Core ML, Core AI, MLX, and Foundation Models.
- The checklist starts from OS version and model job.
- The three sibling courses and the three source links are present.
- Progressive `.flow` diagrams and an Agent brief link are present.