Yes — with 3.9 GB to spare

Granite 4.1 8B at Q4_K_M fits your M1 · 16 GB entirely in unified memory at 8K context, at an estimated 7.7 tokens per second. Past 33K 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 · 4.9 GB Apache 2.0 Released 29 Apr 2026

Matches the old Granite 4.0 32B MoE at a quarter of the size. Fast, Apache 2.0, no reasoning traces.

What hardware do I need for Granite 4.1 8B? →

The VRAM budget

weights 4.9 GB
Weights 4.9 GB KV cache @ 8K 1.25 GB Runtime overhead 0.6 GB Free 3.9 GB of 10.7 GB

Quantisation ladder

QuantWeightsTotal @ 8KMax contextTok/sQualityFit
F16 16.4 GB 18.2 GB ~2.3 Reference 7.5 GB over
Q8_0 8.7 GB 10.5 GB 8K 4.4 −0.1% ppl Fits
Q6_K 6.7 GB 8.6 GB 21K 5.7 −0.4% ppl Long context
Q5_K_M 5.8 GB 7.7 GB 27K 6.5 −0.8% ppl Long context
Q4_K_M 4.9 GB 6.8 GB 33K 7.7 −1.9% ppl Recommended
Q3_K_M 4.0 GB 5.9 GB 39K 9.5 −5.4% ppl Long context
Q2_K 3.4 GB 5.3 GB 42K 11 −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

terminal
$ pip install mlx-lm
$ mlx_lm.generate --model mlx-community/granite-4.1-8b-4bit \
    --max-tokens 512 --prompt "Hello"

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

01Download is 4.9 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 10.7 GB of 16 GB. Raising it with iogpu.wired_limit_mb is possible, and risky.
03There is room to go to 33K context on this card.
See all models for this rig Compare with another model