Apple's Foundation Models framework includes PromptBuilder and InstructionsBuilder—Swift result builders that transform how we compose prompts. This post explores why declarative prompt composition beats string concatenation, and how to build type-safe, composable AI interactions.

Apple Intelligence supports 23 locales, and Foundation Models supports all of them. This chapter shows how to inspect supported locales at runtime and fetch responses in the user's language for regional differences.

An AI-driven journey in porting Qwen 3VL 4B vision model from MLX Python to Swift, featuring coding sessions with GPT-5 Codex, debugging with Sonnet 4.5, and performance optimizations with Haiku 4.5 to achieve code parity.

Learn how to combine Foundation Models' @Generable macro with MLX Swift to get structured, type-safe responses from your on-device models. Build apps that can reliably extract structured data from language models with guaranteed JSON schema compliance.

Grok Code fast (formerly sonic) tested on 60K-line Zenther app. Blazing speed vs limitations analysis, 432M tokens for $13, and why it's my favorite model of 2025 for iOS development.