Yes — with 16.8 GB to spare
Olmo 3 7B Instruct at Q4_K_M fits your M2 Pro · 32 GB entirely in unified memory at 8K context, at an estimated 27 tokens per second. There is room for its full 64K window.
Fully in unified memory
8K context
Q4_K_M · 4.1 GB
Apache 2.0
Released 20 Nov 2025
Fully open — training data and code included. Full multi-head attention, so its cache is 4× a GQA 7B at the same context.
The VRAM budget
weights 4.1 GB
Weights 4.1 GB
KV cache @ 8K 2.50 GB
Runtime overhead 0.6 GB
Free 16.8 GB of 24.0 GB
Quantisation ladder
| Quant | Weights | Total @ 8K | Max context | Tok/s | Quality | Fit |
|---|---|---|---|---|---|---|
| F16 | 13.6 GB | 16.7 GB | 64K | 8.2 | Reference | Long context |
| Q8_0 | 7.2 GB | 10.3 GB | 64K | 15 | −0.1% ppl | Long context |
| Q6_K | 5.6 GB | 8.7 GB | 64K | 20 | −0.4% ppl | Long context |
| Q5_K_M | 4.8 GB | 7.9 GB | 64K | 23 | −0.8% ppl | Long context |
| Q4_K_M | 4.1 GB | 7.2 GB | 64K | 27 | −1.9% ppl | Recommended |
| Q3_K_M | 3.3 GB | 6.4 GB | 64K | 34 | −5.4% ppl | Long context |
| Q2_K | 2.8 GB | 5.9 GB | 64K | 39 | −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. This model interleaves sliding-window layers (4096 tokens, 8 of 32 layers global), which is why its cache barely grows with context.
How to run it
$ pip install mlx-lm $ mlx_lm.generate --model mlx-community/Olmo-3-7B-Instruct-4bit \ --max-tokens 512 --prompt "Hello"
Apple's own array framework. The fastest path on Apple Silicon. More on MLX.
01Download is 4.1 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 the model's full 64K context on this card.