Realtime AI News
oMLX Creator Jun Kim Joins Hugging Face to Support the MLX Community
Hugging Face said in a blog post on September 22 that Jun Kim, the creator and maintainer of oMLX, has joined the company to support the MLX community. The move links Apple's MLX ecosystem more closely to the Hugging Face open-source stack and signals more upstream attention for running models locally on Apple Silicon.
Hugging Face said in a blog post on September 22 that Jun Kim, the creator and maintainer of oMLX, has joined the company, with the stated aim of supporting the MLX community.
Beyond the personnel line, the post reads as a statement about where the platform intends to invest. MLX is Apple's open-source machine learning framework for Apple Silicon, and oMLX is one of the community projects that grew up around running models locally on Macs.
Bringing a project's lead maintainer in-house is one of the most direct ways a platform company can shape an ecosystem. It concentrates knowledge of the project inside the vendor, gives the maintainer paid time to work on it, and tends to move the surrounding toolchain closer to first-class support.
MLX matters because Apple Silicon has become a serious target for local inference. Developers who want to run models on a Mac laptop instead of rented GPUs depend on a small number of projects, and the health of that layer decides how quickly new models become usable on Apple hardware.
For the community, the open question is governance. When a maintainer is employed by a platform company, contributors usually get faster reviews and steadier releases, but they also have to trust that roadmap decisions serve the ecosystem rather than one vendor's priorities.
What to watch over the next few releases: whether MLX support surfaces more visibly across Hugging Face's tooling and model hub, whether oMLX keeps an independent release cadence, and whether maintainers of other Apple-Silicon-adjacent projects get similar offers.
Why it matters
Platform-level backing should make local inference on Apple Silicon smoother for developers, but it also pulls oMLX's governance toward Hugging Face, so the community will watch whether the project keeps an independent cadence.
Nearby Updates
All09/22, 07:40
Inspur launches Yuanbrain SD200 Ultra, claiming one machine can host 2.8-trillion-parameter Kimi K3
Inspur has released the Yuanbrain SD200 Ultra, claiming a single machine can host Kimi K3, a model with 2.8 trillion parameters. The pitch moves very large model deployment from rack-scale clusters toward one box, though memory, interconnect, throughput, price and availability details were not disclosed.
09/22, 07:00
Spain's privacy regulator investigates an AI agent-driven cyber attack
Spain's privacy regulator is investigating a cyber attack carried out with the help of an AI agent, according to a report by teiss. The case raises the question of how data protection rules apply when autonomous software, rather than a human operator, drives the intrusion.
09/22, 09:39
Report: OpenAI and Anthropic pursued an agreement to mutually verify AI models
A report from the Korean broadcaster SBS says OpenAI and Anthropic pursued an agreement to mutually verify each other's AI models. Two of the most direct frontier competitors looking at ways to check a rival's systems points to a shared worry about how credible safety claims really are.
09/22, 10:03
Light Origins open-sources Light-O1-Preview, a 6B whole-body motion model for robots
Light Origins has open-sourced Light-O1-Preview, a roughly 6-billion-parameter whole-body action model released under Apache 2.0 on Hugging Face and GitHub. It pretrains on action priors recovered from human video and transfers the resulting motion prior across robot bodies such as LightBot and the Unitree G1.