Mac LLM
Run LLMs on a Mac
Check what fits your Mac, then run an LLM.
Apple silicon, next to NVIDIA CUDA and AMD.
Apple silicon shares one memory pool between the CPU and GPU.
Only part of that pool is usable for a model.
The fit page estimates memory for your Mac and your context length.
Macs are new, and we're still testing on real hardware.
Questions
Can my Mac run this model?
Enter your memory on the fit page. It shows estimates, not measurements.
Which Macs work?
Apple silicon, M1 or newer. Intel Macs are not supported.