[tumetab1]

Not yet, tried Gemma 4 on an Apple M4 but the tok/s is significant lower than the cloud offering.

Also,the lack of enterprise tooling to help selected an appropriate model and tooling to run a local LLM does not help.