oMLX

oMLX is a macOS application for running large language models locally on Apple Silicon hardware. It uses Apple’s MLX framework for efficient neural network computation, providing a native Mac experience for local LLM inference.

Overview

oMLX is built specifically for Apple Silicon (M-series chips) and leverages Apple’s MLX framework — an array framework designed for machine learning research on Apple hardware. It provides a native Swift/Objective-C interface for loading and running models, with support for various model architectures.

Key Features

  • Apple Silicon Native: Optimized for M1/M2/M3/M4 chips using Apple’s MLX framework
  • Local Inference: Run models entirely offline on your Mac
  • MLX Backend: Uses Apple’s unified memory architecture for efficient CPU/GPU computation
  • Model Support: Compatible with models converted to MLX format
  • Research-Oriented: Targets ML researchers and developers who want to experiment with LLMs on Apple hardware
  • Open Source: The underlying MLX framework is open source (MIT License)

Licensing

oMLX uses the MIT License. Both the oMLX application and Apple’s underlying MLX framework are open source and free to use.

See MLX GitHub for license details.

Official Resources