Yes — with 44.7 GB to spare
Llama 3.2 3B Instruct at Q4_K_M fits your M4 Max · 64 GB entirely in unified memory at 8K context, at an estimated 127 tokens per second. There is room for its full 128K window.
Fully in unified memory
8K context
Q4_K_M · 1.8 GB
Llama 3.2 Community
Released Sep 2024
The 2024 "it just runs" model for 8 GB laptops. Qwen3.5 4B does the same job better now.
The VRAM budget
weights 1.8 GB
Weights 1.8 GB
KV cache @ 8K 0.88 GB
Runtime overhead 0.6 GB
Free 44.7 GB of 48.0 GB
Quantisation ladder
| Quant | Weights | Total @ 8K | Max context | Tok/s | Quality | Fit |
|---|---|---|---|---|---|---|
| F16 | 6.0 GB | 7.5 GB | 128K | 38 | Reference | Long context |
| Q8_0 | 3.2 GB | 4.7 GB | 128K | 72 | −0.1% ppl | Long context |
| Q6_K | 2.5 GB | 3.9 GB | 128K | 93 | −0.4% ppl | Long context |
| Q5_K_M | 2.1 GB | 3.6 GB | 128K | 108 | −0.8% ppl | Long context |
| Q4_K_M | 1.8 GB | 3.3 GB | 128K | 127 | −1.9% ppl | Recommended |
| Q3_K_M | 1.5 GB | 2.9 GB | 128K | 157 | −5.4% ppl | Long context |
| Q2_K | 1.3 GB | 2.7 GB | 128K | 183 | −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/Llama-3.2-3B-Instruct-4bit \ --max-tokens 512 --prompt "Hello"
Apple's own array framework. The fastest path on Apple Silicon. More on MLX.
01Download is 1.8 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 48.0 GB of 64 GB. Raising it with iogpu.wired_limit_mb is possible, and risky.
03There is room to go to the model's full 128K context on this card.