Local Multimodal LLM on iOS with `llama.cpp` (Swift + ObjC++)

This article describes how to create a local multimodal LLM on iOS using llama.cpp. The author provides a step-by-step tutorial on how to create a Swift API that interacts with the llama.cpp library, which is optimized for Metal/ANE. The goal is to create a real-time image processing pipeline with no cloud dependency. The author uses a combination of Swift, Objective-C++, and XCFramework to achieve this.