I'm thinking to buy one. Just need to understand what I'm getting into before I do. My main question is - how does it handle large models? I'm talking about 100-150gb MLX models. How's the speed? And what context? Is it workable for agentic coding? Would appreciate honest answers. Thank you!
