What can Apple Mac Studio (M3 Ultra) run locally?

256 models fit comfortably on Apple Mac Studio (M3 Ultra) with 256 GB system RAM. 0 more run with tradeoffs. Top pick: Hermes 3 Llama 3.1 8B.

What's the recommended runner for this hardware?

MLX-LM (Apple Metal). This is the runner that gives best tokens-per-second on Apple Mac Studio (M3 Ultra) based on macos support and the GPU backend.

What can Apple Mac Studio (M3 Ultra) run for creative?

Build: Mac Studio M3 Ultra 256GB

Memory: 256 GB unified memory

Runner: MLX-LM (Apple Metal)

Any Chat Coding Agents Reasoning Vision Long context Creative

Runs comfortably
256 models

Ranked by fit for creative use case + predicted speed. Click a row for VRAM breakdown.

#1Hermes 3 Llama 3.1 8B

hermes

Commercial OK

Quant: Q8_0Context: 8,192VRAM: 13.4 GBHeadroom: 234.6 GB

ollama run hermes3:8b

tok/s

Estimated

Weights

8.50 GB

KV cache

4.00 GB

Activations

0.43 GB

Runtime

0.50 GB

Model details →

#2Gemma 2 9B Instruct

gemma

Commercial OK

Quant: Q8_0Context: 8,192VRAM: 15.3 GBHeadroom: 232.7 GB

ollama run gemma2:9b

tok/s

Estimated

Weights

9.80 GB

KV cache

4.50 GB

Activations

0.50 GB

Runtime

0.50 GB

Model details →

#3Hermes 3 Llama 3.1 70B

70B

hermes

Commercial OK

Quant: Q4_K_MContext: 8,192VRAM: 77.5 GBHeadroom: 170.5 GB

ollama run hermes3:70b

tok/s

Estimated

Weights

40.00 GB

KV cache

35.00 GB

Activations

2.01 GB

Runtime

0.50 GB

Model details →

#4Hermes 3 Llama 3.2 3B

hermes

Commercial OK

Quant: Q4_K_MContext: 8,192VRAM: 3.9 GBHeadroom: 244.1 GB

243

tok/s

Estimated

Weights

1.80 GB

KV cache

1.50 GB

Activations

0.10 GB

Runtime

0.50 GB

Model details →

#5Dolphin 3.0 Llama 3.2 3B

dolphin

Commercial OK

Quant: Q4_K_MContext: 8,192VRAM: 3.9 GBHeadroom: 244.1 GB

243

tok/s

Estimated

Weights

1.80 GB

KV cache

1.50 GB

Activations

0.10 GB

Runtime

0.50 GB

Model details →

#6Dolphin 3.0 8B

dolphin

Commercial OK

Quant: Q4_K_MContext: 8,192VRAM: 9.8 GBHeadroom: 238.2 GB

ollama run dolphin3:8b

tok/s

Estimated

Weights

5.00 GB

KV cache

4.00 GB

Activations

0.26 GB

Runtime

0.50 GB

Model details →

#7Gemma 2 2B Instruct

gemma

Commercial OK

Quant: Q4_K_MContext: 8,192VRAM: 2.7 GBHeadroom: 245.3 GB

364

tok/s

Estimated

Weights

1.10 GB

KV cache

1.00 GB

Activations

0.06 GB

Runtime

0.50 GB

Model details →

#8Gemma 4 E2B (Effective 2B)

gemma

Commercial OK

Quant: Q8_0Context: 8,192VRAM: 3.8 GBHeadroom: 244.2 GB

ollama run gemma4:e2b

207

tok/s

Estimated

Weights

2.20 GB

KV cache

1.00 GB

Activations

0.12 GB

Runtime

0.50 GB

Model details →

#9CodeGemma 7B

gemma

Commercial OK

Quant: Q4_K_MContext: 8,192VRAM: 8.4 GBHeadroom: 239.6 GB

ollama run codegemma:7b

104

tok/s

Estimated

Weights

4.20 GB

KV cache

3.50 GB

Activations

0.22 GB

Runtime

0.50 GB

Model details →

#10Gemma 4 E4B (Effective 4B)

gemma

Commercial OK

Quant: Q8_0Context: 8,192VRAM: 7.1 GBHeadroom: 240.9 GB

ollama run gemma4:e4b

104

tok/s

Estimated

Weights

4.40 GB

KV cache

2.00 GB

Activations

0.23 GB

Runtime

0.50 GB

Model details →

#11Gemma 3 4B

gemma

Commercial OK

Quant: Q8_0Context: 8,192VRAM: 7.1 GBHeadroom: 240.9 GB

ollama run gemma3:4b

104

tok/s

Estimated

Weights

4.40 GB

KV cache

2.00 GB

Activations

0.23 GB

Runtime

0.50 GB

Model details →

#12Llama 3.2 3B Instruct

llama

Commercial OK

Quant: Q8_0Context: 8,192VRAM: 5.6 GBHeadroom: 242.4 GB

ollama run llama3.2:3b

138

tok/s

Estimated

Weights

3.40 GB

KV cache

1.50 GB

Activations

0.18 GB

Runtime

0.50 GB

Model details →

What if you upgraded?

Hypothetical scenarios. We re-ran the compatibility engine for each.

Move up an Apple memory tier

~$200–400 over base

On Apple Silicon, more unified memory is the only path forward — VRAM and system RAM are the same pool.

Shop this upgrade↗

Some links above are affiliate links. We may earn a commission at no extra cost to you. How we make money.

Won't run
top 5 popular models

Need more memory than you have. Shown for orientation.

DeepSeek V4 Pro (1.6T MoE)

1600B

deepseek

Commercial OK

Needs ~1024 GB unified memory minimum at smallest quant; you have 248 GB available after OS overhead.

—

Qwen 3.5 235B-A17B (MoE)

397B

qwen

Commercial OK

Needs ~256 GB unified memory minimum at smallest quant; you have 248 GB available after OS overhead.

—

DeepSeek R1 (671B reasoning)

671B

deepseek

Commercial OK

Needs ~420 GB unified memory minimum at smallest quant; you have 248 GB available after OS overhead.

—

GLM-5.2

753B

glm

Commercial OK

Needs ~528 GB unified memory minimum at smallest quant; you have 248 GB available after OS overhead.

—

Mistral Medium 3.5 (675B MoE)

675B

mistral

Needs ~448 GB unified memory minimum at smallest quant; you have 248 GB available after OS overhead.

—

How to read these numbers

Measured here

Measured here - RunLocalAI ran this exact combo on owner hardware with public evidence.

Source-backed

Source-backed / community - a reproduced public source supports the speed, but it is not labeled as owner-measured.

Extrapolated

Extrapolated - predicted from a measured benchmark on similar-bandwidth hardware.

Estimated

Estimated - formula based on VRAM bandwidth and model architecture; not a benchmark row.

RunLocalAI Will-It-Run Framework →

Want a specific benchmark we don't have? Email Contact support and we'll prioritize it.

Runs comfortably256 models

What if you upgraded?

Move up an Apple memory tier

Won't runtop 5 popular models

How to read these numbers

Runs comfortably
256 models

Won't run
top 5 popular models