High-performance MLX-based LLM inference engine for macOS with native Swift implementation - View it on GitHub