It's based off of alpaca.cpp, and llama.cpp recently added support for it. So yes, it's Apple Silicon ready.