Hugging Face has brought on Jun Kim, creator of the open-source oMLX framework, as a full-time team member to lead development of local AI tools optimized for Apple Silicon. oMLX, which Hugging Face has hosted since 2023, will maintain its Apache 2.0 license while receiving dedicated resources to accelerate development and strengthen integration with related projects across the machine learning ecosystem. The hire underscores Hugging Face's strategic focus on local AI inference—running models directly on edge devices rather than relying on cloud servers. oMLX is specifically optimized for Apple's hardware, making it critical infrastructure as more developers seek privacy-preserving, offline AI capabilities. The move allows Kim to transition from maintaining oMLX as a side project to full-time development with expanded community support. With Kim's increased focus, Hugging Face plans to streamline the transition from standard transformer model definitions to optimized MLX implementations, enabling developers to more quickly port models to Apple devices. The company also signaled closer collaboration with complementary projects like mlx-lm, mlx-vlm, and LMStudio to create a cohesive ecosystem for local AI development on Apple platforms.