AppleMac / unified memoryCurrent

Apple M3 Ultra for local LLMs

Name: LLM Configurator — GPU VRAM Checker
Author: LLM Configurator

Written by Jakub Rusinowski · Last updated July 12, 2026

With 512 GB at 819 GB/s, the M3 Ultra runs 159 catalogued models at Q4_K_M with 8K context. The largest that fits is Nemotron 3 Ultra 550B (~339.3 GB), and the top pick is Beam 1.0 501B-A23B at 15–30 tok/s.

Apple's highest-end Mac Studio configuration launched with up to 512 GB unified memory at 819 GB/s — enough for 70B+ and some 100B+ class models at Q4/Q5. Note: in March 2026 Apple lowered the maximum purchasable configuration to 256 GB due to DRAM shortages; 512 GB units already in the field retain full capability.

Check my hardware →Full benchmark table

Models that run on the M3 Ultra

Q4_K_M, 8K context, 384 GB usable. Ranked by quality and speed.

Model	VRAM	Speed	Fit
Beam 1.0 501B-A23B Reflection Beam	303.5 GB	15–30 tok/s	Fits
Qwen3-Coder 480B-A35B (MoE) Qwen3-Coder	291.6 GB	12–26 tok/s	Fits
MiniMax M3 428B-A23B MiniMax M3	259.2 GB	15–31 tok/s	Fits
MiniMax M3.1 Flash 428B-A23B MiniMax M3	259.2 GB	15–31 tok/s	Fits
Llama 4 Maverick 17B Llama 4	243.9 GB	22–46 tok/s	Fits
Trinity-Large-Thinking Trinity	242 GB	30–62 tok/s	Fits
Qwen 3.5 397B-A17B Qwen 3.5	240.5 GB	18–38 tok/s	Fits
Nex-N2.5 Pro Nex-N2.5	240.5 GB	18–38 tok/s	Fits
Nex-N2 Pro Nex-N2	240.5 GB	18–38 tok/s	Fits
K2 Horizon 375B-A23B K2 Horizon	231.8 GB	17–35 tok/s	Fits
GLM-5.3-Flash 320B-A18B GLM-5.3-Flash	194 GB	18–37 tok/s	Fits
MiMo-V2.5 310B MiMo-V2.5	188 GB	20–42 tok/s	Fits

Showing 12 of 159

Buy it or rent the same memory

Buy the card, or rent a GPU with the same memory by the hour to try models first.

Buy this hardware Apple Mac Studio M3 Ultra — 512 GB VRAM · 60 W board power Deploy in the cloud now RunPod

or compare on Vast.ai

As an Amazon Associate we earn from qualifying purchases. Cloud GPU links are referral links — we may earn a commission at no extra cost to you.

Affiliate disclosure: Some links on this page are affiliate links — if you buy through them, LLM Configurator may earn a commission at no extra cost to you. As an Amazon Associate, LLM Configurator earns from qualifying purchases.

Apple Mac Studio M3 Ultra

512 GB VRAM · 60 W board power

2026 prices are volatile — check the current listing.

Check price on Amazon

Speed vs other GPUs

Llama 3.1 8B, Q4_K_M. Estimated ranges. How this is calculated

M5 Ultra59–123 tok/s M3 Ultra43–90 tok/s M1 Ultra43–89 tok/s M2 Ultra43–89 tok/s M5 Max34–71 tok/s M4 Max31–64 tok/s M5 Max (32-core GPU)27–56 tok/s M4 Max (32-core GPU)24–50 tok/s

Specifications

Specs last updated 2026-07-12.

Memory: 512 GB
Memory bandwidth: 819 GB/s
Architecture: ARM, 3nm TSMC
Series: Apple Silicon
Board power: 60 W
Release year: 2025
Usable for models: 384 GB (~75% of unified)

Best for70B+ modelslargest local LLMsMac Studioresearch

Recommended first model (Ollama)

Frequently asked questions

Can the Apple M3 Ultra run local LLMs?

Yes. With 512 GB (384 GB usable by a model) it runs 159 of the catalogued models at Q4_K_M with 8K context; the largest is Nemotron 3 Ultra 550B, needing about 339.3 GB.

How fast is the Apple M3 Ultra for AI inference?

It is estimated to run Llama 3.1 8B at 43–90 tok/s at Q4_K_M. Llama 3.3 70B is estimated at 6.5–13 tok/s. These are modelled estimates from memory bandwidth, not measurements; the methodology page shows the formula.

What LLMs can I run on 512 GB?

Among the best that fit: Beam 1.0 501B-A23B, Qwen3-Coder 480B-A35B (MoE), MiniMax M3 428B-A23B, MiniMax M3.1 Flash 428B-A23B, Llama 4 Maverick 17B. The quickest start is Ollama: ollama run llama3.3:70b.

Can I run it on the M3 Ultra?

More for this card

Tools

Guides