Tiny-LLM is an exciting development for implementing large language models (LLMs) on Apple Silicon, specifically aimed at systems engineers. Users are particularly praising the MLX framework, which has rapidly gained traction in the last year for optimizing the performance of LLMs on macOS. The performance metrics shared highlight significant efficiency in terms of tokens processed per second on both power and battery, showcasing the capabilities of Apple’s M1 and M4 Max chips compared to traditional NVIDIA models. As the MLX community grows, there is a call for further exploration of power tuning to maximize LLM performance. Additionally, there’s interest from users in understanding the compatibilities of these models with various iPhones, indicating a broader trend towards mobile and energy-efficient AI processing.