Menu
mutinai
What can I run?
Public preview · figures are illustrative fixture data, attributed to their source · data to May 2025 · read-only

Mac mini M4 Pro 48 GB

What can I run? · 8K context
Change hardwareCurrently: Mac mini M4 Pro 48 GB

Laptops

MacBook Pro M3 Max 64 GBMacBook Pro M4 Max 128 GB

Mini PCs

GB10 mini workstationMac mini M4 Pro 48 GBRyzen AI Max+ 395 mini PC

Desktops

CPU-only Ryzen 9 7950XDual RTX 3090Mac Studio M2 Ultra 192 GBRTX 4060 Ti 16GB budget buildRTX 4090 workstationRTX 5090 workstationRX 7900 XTX desktop

Servers

A100 80GB server
Build your own setupChoose GPUs or a chip and how much memory you have
16run well
0with offload
3too large

Speeds marked measured come from reference results and verified community runs on this exact system; others are estimates.

Runs in accelerator memory 16

Fits entirely in GPU or unified memory — the fast path.

ModelRecommended downloadMemoryGeneration speed
Variant Mixtral 8x7B Instruct v0.1Mistral AI · 46.7B (12.9B active)Q4_K_M via llama.cpp26.6 GiB · LM Studio Community
29.1 of 36.0 GiB
~21 tok/s est.
Variant Qwen2.5 32B InstructQwen Team (Alibaba Cloud) · 32.8B
2 other options
MLX 4-bit · MLX CommunityMLX-LMRuns well20.4 GiB~9 tok/s est.
Q4_K_M · Qwen Team (Alibaba Cloud)llama.cppRuns well21.9 GiB~9 tok/s est.
Q5_K_M via llama.cpp21.7 GiB · LM Studio Community
25.1 of 36.0 GiB
~7 tok/s est.
Variant Qwen2.5-Coder 32B InstructQwen Team (Alibaba Cloud) · 32.8B
1 other option
MLX 4-bit · MLX CommunityMLX-LMRuns well20.4 GiB~9 tok/s est.
Q4_K_M via llama.cpp18.7 GiB · Qwen Team (Alibaba Cloud)
21.9 of 36.0 GiB
~9 tok/s est.
Variant DeepSeek-R1-Distill-Qwen-32BQwen Team (Alibaba Cloud) · 32.8B
1 other option
MLX 4-bit · MLX CommunityMLX-LMRuns well20.4 GiB~9 tok/s est.
Q4_K_M via llama.cpp18.7 GiB · LM Studio Community
21.9 of 36.0 GiB
~9 tok/s est.
Variant Gemma 3 27B ITGoogle DeepMind · 27.4B
1 other option
MLX 4-bit · MLX CommunityMLX-LMRuns well19.3 GiB~11 tok/s est.
Q4_K_M via llama.cpp15.6 GiB · LM Studio Community
20.6 of 36.0 GiB
~10 tok/s est.
Variant Gemma 2 27B ITGoogle DeepMind · 27.2BQ4_K_M via llama.cpp15.5 GiB · LM Studio Community
19.5 of 36.0 GiB
~10 tok/s est.
Variant Mistral Small 24B Instruct 2501Mistral AI · 23.6B
2 other options
MLX 4-bit · MLX CommunityMLX-LMRuns well14.6 GiB~13 tok/s est.
Q4_K_M · LM Studio Communityllama.cppRuns well15.7 GiB~12 tok/s est.
Q6_K via llama.cpp18.0 GiB · LM Studio Community
20.5 of 36.0 GiB
~9 tok/s est.
Variant Qwen2.5 14B InstructQwen Team (Alibaba Cloud) · 14.8B
2 other options
MLX 4-bit · MLX CommunityMLX-LMRuns well10.1 GiB~20 tok/s est.
Q4_K_M · Qwen Team (Alibaba Cloud)llama.cppRuns well10.7 GiB~19 tok/s est.
Q6_K via llama.cpp11.3 GiB · LM Studio Community
13.7 of 36.0 GiB
~14 tok/s est.
Variant Phi-4Microsoft · 14.7B
2 other options
MLX 4-bit · MLX CommunityMLX-LMRuns well10.1 GiB~20 tok/s est.
Q4_K_M · LM Studio Communityllama.cppRuns well10.7 GiB~19 tok/s est.
Q8_0 via llama.cpp14.5 GiB · LM Studio Community
17.1 of 36.0 GiB
~11 tok/s est.
Variant Gemma 2 9B ITGoogle DeepMind · 9.24BQ4_K_M via llama.cpp5.3 GiB · LM Studio Community
8.6 of 36.0 GiB
~29 tok/s est.
Variant Hermes 3 Llama 3.1 8BMeta · 8.03B
2 other options
Q4_K_M · NousResearchllama.cppRuns well6.3 GiB~33 tok/s est.
Q6_K · NousResearchllama.cppRuns well7.9 GiB~25 tok/s est.
Q8_0 via llama.cpp8.0 GiB · NousResearch
9.8 of 36.0 GiB
~20 tok/s est.
Variant Llama 3.1 8B InstructMeta · 8.03B
2 other options
MLX 4-bit · MLX CommunityMLX-LMRuns well5.9 GiB~35 tok/s est.
Q4_K_M · LM Studio Communityllama.cppRuns well6.3 GiB~33 tok/s est.
Q8_0 via llama.cpp7.9 GiB · LM Studio Community
9.8 of 36.0 GiB
~20 tok/s est.
Variant Qwen2.5 7B InstructQwen Team (Alibaba Cloud) · 7.62B
2 other options
MLX 4-bit · MLX CommunityMLX-LMRuns well5.1 GiB~37 tok/s est.
Q4_K_M · Qwen Team (Alibaba Cloud)llama.cppRuns well5.5 GiB~34 tok/s est.
Q8_0 via llama.cpp7.5 GiB · Qwen Team (Alibaba Cloud)
8.8 of 36.0 GiB
~21 tok/s est.
Variant DeepSeek-R1-Distill-Qwen-7BQwen Team (Alibaba Cloud) · 7.62BQ4_K_M via llama.cpp4.3 GiB · LM Studio Community
5.2 of 36.0 GiB
~34 tok/s est.
Variant Mistral 7B Instruct v0.3Mistral AI · 7.25B
1 other option
Q4_K_M · LM Studio Communityllama.cppRuns well5.8 GiB~36 tok/s est.
Q8_0 via llama.cpp7.2 GiB · LM Studio Community
9.0 of 36.0 GiB
~22 tok/s est.
Variant Qwen3 30B-A3BQwen Team (Alibaba Cloud) · 30.5B (3.3B active)
5 other options
MLX 4-bit · MLX CommunityMLX-LMRuns well17.9 GiB~74 tok/s est.
Q4_K_M · Qwen Team (Alibaba Cloud)llama.cppRuns well19.2 GiB~70 tok/s est.
Q4_K_M · Unsloth AIllama.cppRuns well19.3 GiB~70 tok/s est.
MLX 8-bit · MLX CommunityMLX-LMTight fit32.7 GiB~44 tok/s est.
Q8_0 · Qwen Team (Alibaba Cloud)llama.cppTight fit32.7 GiB~44 tok/s est.
Q8_0 via llama.cpp30.2 GiB · LM Studio CommunityTight fit
32.7 of 36.0 GiB
~44 tok/s est.
Too large for this hardware3 model variants
How this is calculatedMemory, fit and speed estimates

Memory = download size + fp16 KV cache for 8K tokens + runtime overhead. Usable memory is 95% of dedicated VRAM, a device-specific share of unified memory (75% by default), and 80% of system RAM for runtimes that can offload. Runtimes must load the file format and support a backend present on the hardware.

Estimated speed is bounded by memory bandwidth ÷ bytes read per token (active parameters for mixture-of-experts), shown in italics with est. and a ±35% range. Measured speeds are medians of reference results and verified public community runs on the same system, download and runtime. For each model we recommend the highest-precision download that fits, preferring not to offload.