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

Build: Apple Mac Studio (M3 Ultra) + — + 192 GB RAM (macos)

Memory: 192 GB unified memory
Runner: MLX-LM (Apple Metal)
AnyChatCodingAgentsReasoningVisionLong contextCreative

Runs comfortably
213 models

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

#1Hermes 3 Llama 3.1 8B
8B
hermes
Commercial OK
Quant: Q8_0Context: 8,192VRAM: 13.4 GBHeadroom: 170.6 GB
ollama run hermes3:8b
52
tok/s
Estimated
Weights
8.50 GB
KV cache
4.00 GB
Activations
0.43 GB
Runtime
0.50 GB
Model details →
#2Dolphin 3.0 Mistral 24B
24B
dolphin
Commercial OK
Quant: Q4_K_MContext: 8,192VRAM: 27.2 GBHeadroom: 156.8 GB
ollama run dolphin-mistral:24b
30
tok/s
Estimated
Weights
14.00 GB
KV cache
12.00 GB
Activations
0.71 GB
Runtime
0.50 GB
Model details →
#3Qwen 2.5 7B Instruct
7B
qwen
Commercial OK
Quant: Q8_0Context: 8,192VRAM: 9.5 GBHeadroom: 174.5 GB
ollama run qwen2.5:7b
59
tok/s
Estimated
Weights
8.10 GB
KV cache
0.47 GB
Activations
0.41 GB
Runtime
0.50 GB
Model details →
#4Qwen 2.5 14B Instruct
14B
qwen
Commercial OK
Quant: Q8_0Context: 8,192VRAM: 24.0 GBHeadroom: 160.0 GB
ollama run qwen2.5:14b
30
tok/s
Estimated
Weights
15.70 GB
KV cache
7.00 GB
Activations
0.79 GB
Runtime
0.50 GB
Model details →
#5Qwen3 Coder 30B-A3B
30B
qwen
Commercial OK
Quant: Q4_K_MContext: 8,192VRAM: 35.5 GBHeadroom: 148.5 GB
ollama run qwen3-coder:30b
24
tok/s
Estimated
Weights
19.00 GB
KV cache
15.00 GB
Activations
0.96 GB
Runtime
0.50 GB
Model details →
#6Mistral 7B Instruct v0.3
7B
mistral
Commercial OK
Quant: Q5_K_MContext: 8,192VRAM: 9.4 GBHeadroom: 174.6 GB
ollama run mistral:7b
91
tok/s
Estimated
Weights
5.10 GB
KV cache
3.50 GB
Activations
0.26 GB
Runtime
0.50 GB
Model details →
#7Llama 3.1 Nemotron Nano 8B
8B
llama
Commercial OK
Quant: Q4_K_MContext: 8,192VRAM: 9.7 GBHeadroom: 174.3 GB
91
tok/s
Estimated
Weights
4.90 GB
KV cache
4.00 GB
Activations
0.25 GB
Runtime
0.50 GB
Model details →
#8North Mini Code 1.0
30B
other
Commercial OK
Quant: Q4_K_MContext: 8,192VRAM: 35.5 GBHeadroom: 148.5 GB
ollama run north-mini-code-1.0
24
tok/s
Estimated
Weights
19.00 GB
KV cache
15.00 GB
Activations
0.96 GB
Runtime
0.50 GB
Model details →
#9GLM-4.7-Flash
31B
glm
Commercial OK
Quant: Q4_K_MContext: 8,192VRAM: 36.0 GBHeadroom: 148.0 GB
ollama run glm-4.7-flash
24
tok/s
Estimated
Weights
19.00 GB
KV cache
15.50 GB
Activations
0.96 GB
Runtime
0.50 GB
Model details →
#10Ornith 1.0 9B
9B
other
Commercial OK
Quant: Q4_K_MContext: 8,192VRAM: 10.9 GBHeadroom: 173.1 GB
ollama run ornith:9b
81
tok/s
Estimated
Weights
5.60 GB
KV cache
4.50 GB
Activations
0.29 GB
Runtime
0.50 GB
Model details →
#11Laguna XS 2.1
33B
other
Commercial OK
Quant: Q4_K_MContext: 8,192VRAM: 38.0 GBHeadroom: 146.0 GB
ollama run laguna-xs-2.1
22
tok/s
Estimated
Weights
20.00 GB
KV cache
16.50 GB
Activations
1.01 GB
Runtime
0.50 GB
Model details →
#12Ornith 1.0 35B
35B
other
Commercial OK
Quant: Q4_K_MContext: 8,192VRAM: 40.1 GBHeadroom: 143.9 GB
ollama run ornith:35b
21
tok/s
Estimated
Weights
21.00 GB
KV cache
17.50 GB
Activations
1.06 GB
Runtime
0.50 GB
Model details →

Runs with tradeoffs
1 models

Tight VRAM, partial CPU offload, or context-limited.

Llama 3.1 Nemotron Ultra 253B
253B
llama
Commercial OK
Quant: Q4_K_MContext: 2,048VRAM: 183.3 GBHeadroom: 0.7 GB
  • • Tight VRAM fit — only 0.7 GB headroom left for context growth
3
tok/s
Estimated
Weights
144.00 GB
KV cache
31.63 GB
Activations
7.20 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 184 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 184 GB available after OS overhead.

—
DeepSeek R1 (671B reasoning)
671B
deepseek
Commercial OK

Needs ~420 GB unified memory minimum at smallest quant; you have 184 GB available after OS overhead.

—
DeepSeek V4 Flash (284B MoE)
284B
deepseek
Commercial OK

Needs ~192 GB unified memory minimum at smallest quant; you have 184 GB available after OS overhead.

—
GLM-5.2
753B
glm
Commercial OK

Needs ~528 GB unified memory minimum at smallest quant; you have 184 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.