Yes — with 14.3 GB to spare
Ministral 3 14B at Q4_K_M fits your M2 Pro · 32 GB entirely in unified memory at 8K context, at an estimated 14 tokens per second. Past 99K the KV cache pushes it over — quantise the cache to q8_0, or step down a quantisation, to go longer.
Fully in unified memory
8K context
Q4_K_M · 7.8 GB
Apache 2.0
Released Dec 2025
Vision
The largest Ministral. A 12 GB card runs it at Q4 with a few gigabytes to spare.
The VRAM budget
weights 7.8 GB
Weights 7.8 GB
KV cache @ 8K 1.25 GB
Runtime overhead 0.6 GB
Free 14.3 GB of 24.0 GB
Quantisation ladder
| Quant | Weights | Total @ 8K | Max context | Tok/s | Quality | Fit |
|---|---|---|---|---|---|---|
| F16 | 25.9 GB | 27.7 GB | — | ~4.3 | Reference | 3.7 GB over |
| Q8_0 | 13.8 GB | 15.6 GB | 61K | 8.1 | −0.1% ppl | Long context |
| Q6_K | 10.6 GB | 12.5 GB | 81K | 11 | −0.4% ppl | Long context |
| Q5_K_M | 9.2 GB | 11.0 GB | 91K | 12 | −0.8% ppl | Long context |
| Q4_K_M | 7.8 GB | 9.7 GB | 99K | 14 | −1.9% ppl | Recommended |
| Q3_K_M | 6.3 GB | 8.2 GB | 109K | 18 | −5.4% ppl | Long context |
| Q2_K | 5.4 GB | 7.3 GB | 115K | 21 | −15% ppl | Long context |
Quality is the published perplexity delta against f16 weights. Max context assumes an f16 KV cache; q8_0 roughly doubles it.
How to run it
$ pip install mlx-lm $ mlx_lm.generate --model mlx-community/Ministral-3-14B-Instruct-2512-4bit \ --max-tokens 512 --prompt "Hello"
Apple's own array framework. The fastest path on Apple Silicon. More on MLX.
01Download is 7.8 GB. Keep it on an SSD — a first load off a spinning disk takes minutes.
02macOS caps what the GPU may wire down at about 24.0 GB of 32 GB. Raising it with iogpu.wired_limit_mb is possible, and risky.
03There is room to go to 99K context on this card.