Jun Kim, oMLX creator and maintainer, joins Hugging Face to support the MLX community

Hugging Face announced on September 22, 2026 that Jun Kim, the creator and maintainer of oMLX, is joining the company to work on the MLX ecosystem. In the announcement, the company said it is fully invested in local AI and called MLX a central piece of that work.

What MLX is, and where Hugging Face fits in

MLX is Apple's framework for local AI, optimized above all for Apple Silicon. Hugging Face notes that it has supported MLX since its release in late 2023, and that its Hub is where users find MLX models and publish their own. The company argues that use of open, local AI keeps accelerating, and that a healthy ecosystem is one where people can pick the tools that work for them.

What changes for oMLX

For oMLX, the headline effect is continuity with more resources: the project graduates from a side job to a fully maintained and funded one. According to Hugging Face, that should let Jun guide contributors better and build for the long term.

The licence does not change. oMLX stays Apache 2.0, and Jun Kim keeps leading it as before.

Plans for the wider MLX ecosystem

Hugging Face describes oMLX as a testbed for new ideas that builds on existing infrastructure such as mlx-lm and mlx-vlm. It plans to upstream work wherever that makes sense, and says it is continuing to collaborate with the teams behind mlx-lm, mlx-vlm and LM Studio.

One concrete focus area is shortening the path from a model definition in the transformers library to a reference MLX implementation that different inference engines can consume. Since transformers has become the reference point for model definitions, the goal is to get new models running on MLX faster, while leaving each engine free to focus on its own distinctive features.