Comment by asenna

You should be able to do some local LLM models on the M2 Max 64gb (Qwen 3.8 27b would be a strong model). I might go for that one.