Yes — with 89.3 GB to spare

Ministral 3 8B at Q4_K_M fits your M3 Max · 128 GB entirely in unified memory at 8K context, at an estimated 45 tokens per second. There is room for its full 256K window.

Fully in unified memory 8K context Q4_K_M · 5.0 GB Apache 2.0 Released Dec 2025 Vision

Mistral's 8B with images in. Plain GQA, so budget more KV cache than Qwen3.5 9B at the same context.

What hardware do I need for Ministral 3 8B? →

The VRAM budget

weights 5.0 GB

Weights 5.0 GB KV cache @ 8K 1.06 GB Runtime overhead 0.6 GB Free 89.3 GB of 96.0 GB

Quantisation ladder

Quant	Weights	Total @ 8K	Max context	Tok/s	Quality	Fit
F16	16.6 GB	18.3 GB	256K	13	Reference	Long context
Q8_0	8.8 GB	10.5 GB	256K	25	−0.1% ppl	Long context
Q6_K	6.8 GB	8.5 GB	256K	33	−0.4% ppl	Long context
Q5_K_M	5.9 GB	7.6 GB	256K	38	−0.8% ppl	Long context
Q4_K_M	5.0 GB	6.7 GB	256K	45	−1.9% ppl	Recommended
Q3_K_M	4.1 GB	5.7 GB	256K	55	−5.4% ppl	Long context
Q2_K	3.5 GB	5.1 GB	256K	64	−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

Ollama llama.cpp LM Studio MLX

terminal

$ pip install mlx-lm
$ mlx_lm.generate --model mlx-community/Ministral-3-8B-Instruct-2512-4bit \
    --max-tokens 512 --prompt "Hello"

Apple's own array framework. The fastest path on Apple Silicon. More on MLX.

01Download is 5.0 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 96.0 GB of 128 GB. Raising it with iogpu.wired_limit_mb is possible, and risky.

03There is room to go to the model's full 256K context on this card.

See all models for this rig Compare with another model