Can I Run Qwen 3.5 on Mac Studio (M2 Ultra, 192 GB)?

Written by Jakub Rusinowski · Last updated February 24, 2026

Yes, but it is tight

Yes, but it is tight — Qwen 3.5 122B-A10B at Q8_0 needs about 133.3 GB of the 144 GB usable on Mac Studio (M2 Ultra, 192 GB), leaving only ~10.7 GB before the runtime starts swapping. Expect ~32.8 tok/s (estimated), with room for about 32,768 tokens of context.

Confidence: low · Recommended quantization: Q8_0 · Estimated speed: ~32.8 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

Qwen 3.5 on Mac Studio (M2 Ultra, 192 GB): memory by quantization

Quant	Memory needed	Fits 144 GB?	Max context	Est. speed	Download
F16	247.7 GB	✗ No	—	—	244 GB
Q8_0	133.3 GB	✓ Yes	32K	~32.8 tok/s	129.6 GB
Q6_K	103.7 GB	✓ Yes	64K	~40.1 tok/s	100 GB
Q5_K_M	90.1 GB	✓ Yes	128K	~44.6 tok/s	86.5 GB
Q4_K_M	77.3 GB	✓ Yes	128K	~49.9 tok/s	73.7 GB
Q3_K_M	55.7 GB	✓ Yes	128K	~62.6 tok/s	52 GB
Q2_K	43.8 GB	✓ Yes	128K	~72.7 tok/s	40.1 GB

Which Qwen 3.5 sizes fit

Variant	Needs at Q4_K_M	Fits?	Est. speed
Qwen 3.5 397B-A17B	244.7 GB	✗ Too large	—
Qwen 3.5 122B-A10B	77.3 GB	✓ Fits	~49.9 tok/s
Qwen 3.5 35B-A3B	23.8 GB	✓ Fits	~107.8 tok/s
Qwen 3.5 27B	18.8 GB	✓ Fits	~23.7 tok/s
Qwen 3.5 9B	7.4 GB	✓ Fits	~59.6 tok/s
Qwen 3.5 4B	4.1 GB	✓ Fits	~104.8 tok/s
Qwen 3.5 2B	2.7 GB	✓ Fits	~151.6 tok/s
Qwen 3.5 0.8B	1.8 GB	✓ Fits	~212.1 tok/s

What to watch out for

1 larger variant of Qwen 3.5 does not fit and would need CPU offload or different hardware.
This model does not publish its full attention configuration, so the KV-cache share of these figures is inferred from its parameter count rather than computed exactly.
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Superseded by the M3 Ultra, which is why it is often the better used buy for large-model work.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 inferred from the parameter count because this model does not publish its full attention configuration.

FAQ