Can I Run DeepSeek-R1-Distill-Qwen-32B on Mac Studio (M3 Ultra, 256 GB)?

Superseded model. DeepSeek R1 has been superseded by DeepSeek V4. This page is kept for reference; the newer family is a better starting point. View DeepSeek V4 →

Written by Jakub Rusinowski · Last updated January 20, 2025

These figures are for DeepSeek-R1-Distill-Qwen-32B, a distill of Qwen2.5-32B — not the full DeepSeek R1. The full DeepSeek R1 (671B) needs about 405 GB of weights at Q4_K_M and is a different model.

Yes — comfortably

Yes, comfortably — DeepSeek R1 Distill Qwen 32B at Q8_0 needs about 36.9 GB of the 192 GB usable on Mac Studio (M3 Ultra, 256 GB), leaving ~155.1 GB spare and running at ~12.3 tok/s (estimated), with room for about 131,072 tokens of context.

Confidence: high · Recommended quantization: Q8_0 · Estimated speed: ~12.3 tok/s

Mac Studio (M3 Ultra, 256 GB) — what it gives a model

Usable memory for models	192 GB
Memory bandwidth	819 GB/s
Form factor	Desktop
Operating system	macOS
Memory upgradeable	No — soldered

DeepSeek R1 on Mac Studio (M3 Ultra, 256 GB): memory by quantization

Quant	Memory needed	Fits 192 GB?	Max context	Est. speed	Download
F16	66.9 GB	✓ Yes	128K	~6.8 tok/s	64 GB
Q8_0	36.9 GB	✓ Yes	128K	~12.3 tok/s	34 GB
Q6_K	29.2 GB	✓ Yes	128K	~15.7 tok/s	26.2 GB
Q5_K_M	25.6 GB	✓ Yes	128K	~17.9 tok/s	22.7 GB
Q4_K_M	22.3 GB	✓ Yes	128K	~20.6 tok/s	19.3 GB
Q3_K_M	16.6 GB	✓ Yes	128K	~28 tok/s	13.6 GB
Q2_K	13.5 GB	✓ Yes	128K	~34.7 tok/s	10.5 GB

Which DeepSeek R1 sizes fit

Variant	Needs at Q4_K_M	Fits?	Est. speed
DeepSeek R1 (671B)	406.5 GB	✗ Too large	—
DeepSeek R1 Distill Qwen 32B	22.3 GB	✓ Fits	~20.6 tok/s
DeepSeek R1 Distill Qwen 14B	10.6 GB	✓ Fits	~42.9 tok/s
DeepSeek R1 Distill Llama 8B	6.7 GB	✓ Fits	~66.9 tok/s

What to watch out for

1 larger variant of DeepSeek R1 does not fit and would need CPU offload or different hardware.
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 M3 Ultra 256 GB limitations

The largest single-box unified memory pool available, but at roughly half the memory bandwidth of a high-end discrete GPU — very large models load, then run slowly.
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.
256 GB unified memory at 819 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

Can I run DeepSeek R1 on Mac Studio (M3 Ultra, 256 GB)?

Which quantization of DeepSeek R1 should I use on Mac Studio (M3 Ultra, 256 GB)?

Q8_0 — it needs about 36.9 GB of the 192 GB available, downloads as roughly 34 GB, and runs at an estimated 12.3 tokens/sec with up to 128K of context.

What limits DeepSeek R1 on Mac Studio (M3 Ultra, 256 GB)?

Nothing binding — the model fits with headroom and generates at a usable speed on this hardware.

Can I Run DeepSeek-R1-Distill-Qwen-32B on Mac Studio (M3 Ultra, 256 GB)?

Mac Studio (M3 Ultra, 256 GB) — what it gives a model

DeepSeek R1 on Mac Studio (M3 Ultra, 256 GB): memory by quantization

Which DeepSeek R1 sizes fit

What to watch out for

Mac Studio M3 Ultra 256 GB limitations

Recommended setup

How these numbers are calculated

FAQ

Can I run DeepSeek R1 on Mac Studio (M3 Ultra, 256 GB)?

Which quantization of DeepSeek R1 should I use on Mac Studio (M3 Ultra, 256 GB)?

What limits DeepSeek R1 on Mac Studio (M3 Ultra, 256 GB)?

Which runtime should I use?

Other Models on Mac Studio (M3 Ultra, 256 GB)

DeepSeek R1 on GPUs

What This Model Is Good At

Model & Tools