Can I Run Command R Family on Mac Studio (M2 Ultra, 192 GB)?

Newer alternative. Cohere has not released a newer model in this library. If you are choosing today, Mistral Small 4 is a comparable current option. View Mistral Small 4 →

Written by Jakub Rusinowski · Last updated April 4, 2024

Yes — comfortably

Yes, comfortably — Command R+ (104B) at Q6_K needs about 88.2 GB of the 144 GB usable on Mac Studio (M2 Ultra, 192 GB), leaving ~55.8 GB spare and running at ~5 tok/s (estimated), with room for about 65,536 tokens of context.

Confidence: high · Recommended quantization: Q6_K · Estimated speed: ~5 tok/s

Mac Studio (M2 Ultra, 192 GB) — what it gives a model

Usable memory for models	144 GB
Memory bandwidth	800 GB/s
Form factor	Desktop
Operating system	macOS
Memory upgradeable	No — soldered

Command R Family on Mac Studio (M2 Ultra, 192 GB): memory by quantization

Quant	Memory needed	Fits 144 GB?	Max context	Est. speed	Download
F16	210.9 GB	✗ No	—	—	208 GB
Q8_0	113.4 GB	✓ Yes	64K	~3.9 tok/s	110.5 GB
Q6_K	88.2 GB	✓ Yes	64K	~5 tok/s	85.3 GB
Q5_K_M	76.7 GB	✓ Yes	64K	~5.8 tok/s	73.7 GB
Q4_K_M	65.7 GB	✓ Yes	64K	~6.7 tok/s	62.8 GB
Q3_K_M	47.3 GB	✓ Yes	64K	~9.4 tok/s	44.3 GB
Q2_K	37.1 GB	✓ Yes	64K	~12 tok/s	34.2 GB

Which Command R Family sizes fit

Variant	Needs at Q4_K_M	Fits?	Est. speed
Command R+ (104B)	65.7 GB	✓ Fits	~6.7 tok/s
Command R (35B)	32.7 GB	✓ Fits	~15.8 tok/s

What to watch out for

Memory on this machine is not upgradeable, so the configuration you buy is the ceiling for every model you will ever run on it.

Mac Studio M2 Ultra 192 GB limitations

Superseded by the M3 Ultra, which is why it is often the better used buy for large-model work.
Memory is soldered; the configuration chosen at purchase is permanent.

Recommended setup

Ollama or LM Studio (Metal) — MLX for the fastest Apple-native throughput

How these numbers are calculated

Memory need = quantized weights + KV cache at 8,192 tokens (f16) + 0.8 GB runtime overhead.
192 GB unified memory at 800 GB/s, shared between CPU and GPU.
macOS reserves a share of unified memory for the system, so not all of it is available to a model.
Throughput is a memory-bandwidth roofline estimate, not a measurement. It is labelled "estimated" everywhere it appears.
KV cache is computed from this model's published attention configuration.

FAQ