Now hip to the Mac mini’s local LLM prowess, Apple is back with new and souped-up Mac minis that swap out the older M4 and M4 Pro chips with mightier M6 and M5 Pro chips (confusingly, the M5 Pro processors are the more powerful ones) offering more CPU and GPU cores, higher memory bandwidth (up to 160GB/s), and most importantly, a solid amount of unified RAM, which is the be-all-end-all of speedy AI inference.
But while the newest Mac minis are still much cheaper than AMD Ryzen AI Halo and Nvidia RTX Spark systems (which, arguably, are actually competing with Apple’s stratospherically priced Mac Studio), their prices have spiked compared to their two-year-old M4-packing predecessors. The reason is… well, you know why.
Apple just announced its new M6 and M5 Pro Mac minis, so we haven’t had the chance to benchmark them compared to the older M4 and M4 Pro models, nor have we stacked them up against AMD’s and Nvidia’s local AI workhorses. For its part, Apple is promising an overall 4x local AI boost versus the older Mac mini models, but we haven’t put that claim to the test yet. (There are also some important “gotchas” to consider with local AI on a Mac mini, which I’ll get to in a bit.)
What I can do is tell you the Mac mini system I’d be eyeing as a local AI user, assuming I were game for buying a premium Apple system at the very height of the PC hardware market—which I’m not, and you probably shouldn’t be either. But hey, at least we can browse, right?
For background, I currently own a Mac Mini M4 Pro with 64GB of unified RAM and 512GB of internal storage, with that storage upped by several external SSDs plus a 1GB Thunderbolt NVMe for loading local AI models. (I should’ve sprung for a bigger NVMe drive, but I was being cheap.)
My particular Mac mini cost me roughly $2,000 in late 2024, and while it can handle the current local LLM favorite, Alibaba’s Qwen 3.8 27B, the MLX variant of that model (optimized to run on Apple Silicon) starts to chug once its context window fills to much higher than 40,000 tokens or so, making the Mac mini’s cooling fans spin up like jet engines.