Skip to content

What can the Mac mini M4 Pro 48 GB run?

What fits on the Mac mini M4 Pro 48 GB
Text models: 218Visual models: 18Strongest first
Memory
48 GB unified
Bandwidth
273 GB/s
Maker
Apple

Open the calculator on this machine

Text models that run on it

Text models that run on the Mac mini M4 Pro 48 GB, best first
ModelBest fitContextSpeedStrongest at
Qwen3.8 27B8-bit88K7.6 tok/sChat, Coding, Agents, Docs
Muse Glimmer 30B8-bit128K7.2 tok/sDocs
Qwen3.6 35B A3B6-bit256K59 tok/sChat, Coding, Agents, Docs
Gemma 4 31B8-bit36K6.5 tok/sChat, Coding, Agents, Docs
Gemma 4 26B A4B8-bit256K45 tok/sChat, Coding, Agents, Docs
Qwen3.5 9B8-bit256K23 tok/sChat, Coding, Agents, Docs
GLM-4.7 Flash8-bit52K48 tok/sChat, Coding, Agents
Gemma 4 12B8-bit256K17 tok/sChat, Coding, Docs
Nemotron 3.5 Lightning 30B A3B8-bit256K46 tok/sAgents, Docs
Devstral Small 2 24BFP860K8.7 tok/sCoding
LFM2.5 24B A2B8-bit128K66 tok/sAgents
Mistral Small 3.2 24B8-bit60K8.8 tok/sChat, Coding, Agents
Ministral 3 14B8-bit120K15 tok/sCoding, Agents, Docs
Gemma 4 E4B8-bit128K36 tok/sChat, Coding, Docs
gpt-oss 20BMXFP4128K67 tok/sAgents
Ministral 3 8B8-bit184K24 tok/sDocs
Olmo 3.1 32B6-bit64K8.2 tok/sChat, Coding, Agents, Docs
Llama 3.1 8B8-bit128K26 tok/sChat, Coding, Agents, Docs
Qwen3 14B8-bit40K14 tok/sNot rated
Qwen3 VL 8B Instruct8-bit176K24 tok/sText and images
Qwen3 8B8-bit40K26 tok/sNot rated
Qwen2.5 7B Instruct8-bit32K28 tok/sNot rated
Qwen2.5 VL 7B Instruct8-bit125K26 tok/sText and images
Qwen3 VL 4B Instruct8-bit200K42 tok/sText and images
Qwen3 4B8-bit40K46 tok/sNot rated
Qwen3 32B6-bit32K8.2 tok/sNot rated
Qwen3 4B Instruct 25078-bit208K46 tok/sNot rated
Qwen3.6 27B8-bit88K7.6 tok/sText and images
Qwen2.5 Coder 14B Instruct8-bit32K14 tok/sCoding
Qwen3 Coder 30B A3B Instruct8-bit32K48 tok/sCoding
Qwen3.5 27B8-bit88K7.6 tok/sText and images
Qwen3.5 35B A3B6-bit256K59 tok/sText and images
NVIDIA Nemotron 3 Nano 30B A3B BF168-bit256K46 tok/sNot rated
Qwen2.5 Coder 7B Instruct8-bit32K28 tok/sCoding
Mistral 7B Instruct V0.28-bit32K28 tok/sNot rated
Qwen3 30B A3B8-bit32K48 tok/sNot rated
Qwen3 30B A3B Instruct 25078-bit32K48 tok/sNot rated
Laguna XS.26-bit200K60 tok/sNot rated
OTel 2.0 LLM 31B IT8-bit36K6.5 tok/sNot rated
Qwen2.5 32B Instruct6-bit32K8.2 tok/sNot rated
MiniCPM V 4.7 35B A3B6-bit256K59 tok/sText and images
Qwen2.5 14B Instruct8-bit32K14 tok/sNot rated
Laguna XS 2.16-bit200K60 tok/sNot rated
Diffusiongemma 26B A4B IT8-bit256K43 tok/sText and images
Qwen2.5 VL 32B Instruct6-bit28K8.0 tok/sText and images
Qwen2.5 Coder 32B Instruct6-bit32K8.2 tok/sCoding
Qwen2 VL 7B Instruct8-bit32K26 tok/sText and images
Ornith 1.5 9B8-bit256K23 tok/sNot rated
Qwen3 VL 30B A3B Instruct8-bit28K48 tok/sText and images
Meta Llama 3 8B Instruct Abliterated V38-bit8K26 tok/sNot rated
EXAONE 3.5 7.8B Instruct8-bit32K27 tok/sNot rated
Saiga Llama3 8B8-bit8K26 tok/sNot rated
DeepSeek R1 0528 Qwen3 8B8-bit128K26 tok/sNot rated
Gemma 3 12B IT8-bit128K18 tok/sText and images
Phi 3.5 Mini Instruct8-bit80K49 tok/sNot rated
DeepSeek R1 Distill Qwen 32B6-bit32K8.2 tok/sNot rated
Qwen3 4B Thinking 25078-bit208K46 tok/sNot rated
Gemma 3 27B IT8-bit72K7.8 tok/sText and images
DeepSeek R1 Distill Qwen 7B8-bit128K28 tok/sNot rated
Qwen2 7B Instruct8-bit32K28 tok/sNot rated

Best fit is the highest quality, up to 8-bit, that leaves the use case’s context. Ranked by our use-case scores, then by Hugging Face downloads. Speeds are rough, plus or minus 25%. The top 60 of 218 are shown.

Image and video models

Image and video models that run on the Mac mini M4 Pro 48 GB
ModelBest fitPer outputStrongest at
LTX-2.58-bit13 minVideo
Z-Image Turbo16-bit1.9 minImages
FLUX.1 [dev]16-bit3.9 minImages
Qwen-Image8-bit75 minImages
Qwen-Image 2.116-bit32 minImages
FLUX.3 Klein16-bit2.9 minImages
Wan2.2 I2V A14B8-bit11.0 hVideo
FLUX.2 [klein] 9B16-bit55 sImages
Wan2.1 I2V 14B8-bit11.0 hVideo
Wan 2.1 T2V-14B8-bit13.8 hVideo
Krea 2 Turbo16-bit3.9 minImages
Stable Diffusion 3.5 Large16-bit17 minImages
Qwen Image 25128-bit75 minImages
MOSS-Image16-bit21 minImages
Wan 2.2 TI2V-5B16-bit3.0 hVideo
MOSS Image Base16-bit21 minImages
Wan 2.2 A14B8-bit11.0 hVideo
HunyuanVideo 1.516-bit13.5 hVideo

Only just too big

Questions

What is the best LLM for the Mac mini M4 Pro 48 GB?

Of the open models we track, the strongest that run well on the Mac mini M4 Pro 48 GB are Qwen3.8 27B, Muse Glimmer 30B, Qwen3.6 35B A3B. 218 text models fit it in all.

Calculated from published specs, not benchmarked. All machines