Mac mini local LLM memory: what each size runs

Apple lists 16 GB as the smallest Mac mini and 64 GB as the largest. At the default macOS GPU share, the smallest Mac mini gives the GPU about 11 GiB and the largest gives about 48 GiB. On the largest configuration, the biggest model in our list that runs is Mixtral-8x7B-Instruct-v0.1 at Q6_K (36.9 GiB–38.0 GiB at 8,192 tokens). On the smallest, it is Mistral-Nemo-Instruct-2407 at Q4_K_M.

Chips and memory sizes Apple lists

Mac mini chip options with memory bandwidth and unified memory sizes
ChipGPUMemory bandwidthUnified memory
Apple M6 (12-core CPU)12-core GPU153 GB/s at 16 GB, 170 GB/s at 24 GB and 32 GB16 GB, 24 GB, 32 GB
Apple M5 Pro (15-core CPU, 18-core CPU option)16-core GPU, or 20-core GPU as an option307 GB/s24 GB, 48 GB, 64 GB (Apple prints 48 GB and 64 GB as options for the M5 Pro with the 18-core CPU and 20-core GPU.)

From Apple's Mac mini technical specifications, read 2026-10-18. Check that page before buying: Apple ties some memory sizes to a specific chip configuration. Bandwidth is Apple's figure. This site does not turn it into tokens per second and has no measurement on these machines.

What each memory size runs

GPU budget and model fit for each Mac mini memory size at 8,192 tokens
Unified memoryGPU budgetModels that runTightLargest that runs
16 GB10.7 GiB8 of 223Mistral-Nemo-Instruct-2407, Q4_K_M
24 GB16.0 GiB11 of 221Qwen2.5-14B-Instruct, Q6_K
32 GB21.3 GiB13 of 223Qwen3-30B-A3B, Q4_K_M
48 GB36.0 GiB17 of 220Mixtral-8x7B-Instruct-v0.1, Q4_K_M
64 GB48.0 GiB17 of 223Mixtral-8x7B-Instruct-v0.1, Q6_K

Budget is the default share macOS gives the GPU: 67% up to 32 GB and 75% above (the rule and sources are on each tier page). Verdicts use the catalog's llama.cpp GGUF files, one sequence, fp16 KV cache and 8,192 tokens, over the 22 models in the Mac list. Mixture-of-experts models count at full file size. Sizes without a tier page (36, 256 and 512 GB) use the same rule. These are memory estimates, not runs on a Mac.

Data sources

Related pages