Apple · M2 Ultra · Desktop

Best AI Models for Mac Studio M2 Ultra (128 GB)

Runs 27–31B and 120B MoE models well; dense 70B fits but is slow at 12 tok/s.

128 GB unified124.5 GB for models800 GB/s54 TFLOPS FP162023

Memory

Same chip in

What LLMs can the Mac Studio M2 Ultra run?

#	Model	Speed	Quant · Memory
1	Qwen3.8 27BAlibabaHow to run → Q4_K_M · 17.9 GB	32.1 tok/s	Q4_K_M · 17.9 GB
2	Qwen3.6 27BAlibaba Q4_K_M · 17.9 GB	32.1 tok/s	Q4_K_M · 17.9 GB
3	Gemma 4 31B ITGoogle Q4_K_M · 21.7 GB	27.5 tok/s	Q4_K_M · 21.7 GB
4	Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB	91.7 tok/s	Q4_K_M · 16.5 GB
5	NVIDIA Nemotron 3.5 Lightning 30B A3B BF16NVIDIA Q4_K_M · 19.3 GB	77.0 tok/s	Q4_K_M · 19.3 GB

#	Model	Speed	Quant · Memory
1	Qwen3.8 27BAlibabaHow to run → Q4_K_M · 17.9 GB	32.1 tok/s	Q4_K_M · 17.9 GB
2	Gemma 4 31B ITGoogle Q4_K_M · 21.7 GB	27.5 tok/s	Q4_K_M · 21.7 GB
3	Qwen3.6 35B A3BAlibaba Q4_K_M · 22.0 GB	75.9 tok/s	Q4_K_M · 22.0 GB
4	Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB	91.7 tok/s	Q4_K_M · 16.5 GB
5	NVIDIA Nemotron 3.5 Lightning 30B A3B BF16NVIDIA Q4_K_M · 19.3 GB	77.0 tok/s	Q4_K_M · 19.3 GB

#	Model	Speed	Quant · Memory
1	Qwen3.8 27BAlibabaHow to run → Q4_K_M · 17.9 GB	32.1 tok/s	Q4_K_M · 17.9 GB
2	Qwen3.5 27BAlibaba Q4_K_M · 17.9 GB	32.1 tok/s	Q4_K_M · 17.9 GB
3	Gemma 4 31B ITGoogle Q4_K_M · 21.7 GB	27.5 tok/s	Q4_K_M · 21.7 GB
4	Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB	91.7 tok/s	Q4_K_M · 16.5 GB
5	NVIDIA Nemotron 3.5 Lightning 30B A3B BF16NVIDIA Q4_K_M · 19.3 GB	77.0 tok/s	Q4_K_M · 19.3 GB

#	Model	Speed	Math Score	Quant · Memory
1	Qwen3.8 27BAlibabaHow to run → Q4_K_M · 17.9 GB	32.1 tok/s		Q4_K_M · 17.9 GB
2	Qwen3.6 27BAlibaba Q4_K_M · 17.9 GB	32.1 tok/s		Q4_K_M · 17.9 GB
3	Qwen3.6 35B A3BAlibaba Q4_K_M · 22.0 GB	75.9 tok/s		Q4_K_M · 22.0 GB
4	Gemma 4 31B ITGoogle Q4_K_M · 21.7 GB	27.5 tok/s	overall	Q4_K_M · 21.7 GB
5	NVIDIA Nemotron 3.5 Lightning 30B A3B BF16NVIDIA Q4_K_M · 19.3 GB	77.0 tok/s		Q4_K_M · 19.3 GB

#	Model	Speed ↓	Quant · Memory
1	GPT OSS 20BOpenAIHow to run → Q4_K_M · 13.1 GB	106 tok/s	Q4_K_M · 13.1 GB
2	Gemma 4 26B A4B ITGoogle Q4_K_M · 16.5 GB	91.7 tok/s	Q4_K_M · 16.5 GB
3	Qwen3.5 9BAlibaba Q4_K_M · 6.6 GB	88.1 tok/s	Q4_K_M · 6.6 GB
4	NVIDIA Nemotron 3.5 Lightning 30B A3B BF16NVIDIA Q4_K_M · 19.3 GB	77.0 tok/s	Q4_K_M · 19.3 GB
5	Qwen3.6 35B A3BAlibaba Q4_K_M · 22.0 GB	75.9 tok/s	Q4_K_M · 22.0 GB

#	Model	Speed	Agents Score	Quant · Memory
1	GPT OSS 120BOpenAIHow to run → Q4_K_M · 70.8 GB	74.6 tok/s		Q4_K_M · 70.8 GB
2	Qwen3.8 27BAlibaba Q4_K_M · 17.9 GB	32.1 tok/s	overall	Q4_K_M · 17.9 GB
3	Qwen3.6 27BAlibaba Q4_K_M · 17.9 GB	32.1 tok/s	overall	Q4_K_M · 17.9 GB
4	NVIDIA Nemotron 3.5 Lightning 30B A3B BF16NVIDIA Q4_K_M · 19.3 GB	77.0 tok/s		Q4_K_M · 19.3 GB
5	GLM 4.7 FlashZ.ai Q4_K_M · 19.9 GB	68.4 tok/s		Q4_K_M · 19.9 GB

See all 367 models on the Mac Studio M2 Ultra →Bars = traction for models without a Score yet
Task rankings use that task's benchmarks, topped up from Overall when short; bars mark well-known models with no Score yet.

About the Mac Studio M2 Ultra (128 GB)

The Mac Studio M2 Ultra is the fastest 128 GB machine: it runs 27–31B and 120B MoE models well. macOS keeps about 3.5 GB, leaving 124.5 GB for models. Prompt reading is its weak spot: a 4K-token prompt waits 21 s before the first word on Qwen3.8 27B. M2 Ultra generation, released 2023; the newer M3 Ultra writes 1% faster with 0.8× the compute.

Speed in 128 GB#1800 GB/s

Prompt reading#454 TFLOPS

For models124.5 GBmacOS keeps 3.5 GB

ChipM2 Ultrareleased 2023

Mac Studio M2 Ultra strengths and weaknesses

Strengths

Runs 120B MoE models: GPT OSS 120B at 74.6 tok/s in 70.8 GB
Room for long context: Qwen3.8 27B at Q4 uses 17.9 GB, leaving 106.6 GB free

Weaknesses

Slow prompt reading: A 32K-token prompt on Qwen3.8 27B waits 2.8 min for the first token
No CUDA: vLLM, AWQ, GPTQ and FP8 builds don't run; use GGUF or MLX

Memory for models

124.5 GB of 128 GB

35B MoE22 GB70B51.3 GB123B74.7 GB

35B MoE22 GB123B74.7 GB

For models 124.5 GBmacOS keeps 3.5 GBModel size at Q4

Everything on the size ladder fits. Two 27B-class models together use 35.7 GB and leave 88.8 GB free.

Similar devices and alternatives

Other memory sizes of the same machine first, then machines and GPUs people weigh against it, each with the one reason it is worth a look. Speeds on Qwen3.6 35B A3B.

Same chip, 64 GB less memoryMac Studio M2 Ultra (64 GB)Apple · 60.5 GB for models · M2 Ultra

same speed▼ 64 GBno price

Same family Same chip, 64 GB more memoryMac Studio M2 Ultra (192 GB)Apple · 188.5 GB for models · M2 Ultra

same speed▲ 64 GB$5,599

Same family Newer machine, 15% slowerMac Studio M4 Max (128 GB)Apple · 124.5 GB for models · M4 Max

▼ 15% speedsame memory$3,499

Reads prompts 5× faster, 120 GB for modelsDGX SparkNVIDIA · 120 GB for models · GB10

▲ 5× prompts▼ 29% speed▼ 4.5 GB$3,999

2.0× faster if your models fit in 24 GBRTX 4090NVIDIA · 24 GB VRAM · GPU

▲ 99% speed▼ 100.5 GB$1,599rent $0.34/hr

2.8× faster if your models fit in 32 GBRTX 5090NVIDIA · 32 GB VRAM · GPU

▲ 181% speed▼ 92.5 GB$5,376rent $0.70/hr

Mac Studio M2 Ultra tokens per second by model size

Output speed depends on memory bandwidth (800 GB/s). Grey rows need more than the 124.5 GB available.

9BQwen3.5 9B · Q4_K_M · 6.6 GB
88.1tok/s
14BQwen3 14B · Q4_K_M · 11.9 GB
58.9tok/s
27–31BQwen3.8 27B · Q4_K_M · 17.9 GB
32.1tok/s
35B MoEQwen3.6 35B A3B · Q4_K_M · 22.0 GB
75.9tok/s
70BLlama 3.3 70B Instruct · Q4_K_M · 51.3 GB
12.0tok/s
120B MoEGPT OSS 120B · Q4_K_M · 70.8 GB
74.6tok/s

Estimated speeds: Q4_K_M, 4K context, one user. How we estimate

Ultra generations: writing vs compute

M1 Ultra → M3 Ultra

Writing speed (tok/s)Compute (TFLOPS FP16)

Writing speed follows memory bandwidth and barely moved: 75.9 to 76.6 tok/s since the M1 Ultra. Compute, which sets how fast prompts are read, grew 1.4×. Indexed to the M1 Ultra; speeds on Qwen3.6 35B A3B, same 128 GB.

All 1228 models on the Mac Studio M2 Ultra

Featured only

Best quant that fits in 124.5 GB, ordered by llmrun Score. Fit grades how well it runs here, S best. How ranking works

LLM models compatible with Mac Studio M2 Ultra (128 GB) — ranked by Score, recency and Fit
Model	Quant	VRAM	Speed	Context	Fit
Phi 3.5 Mini Instruct3.8BChatCode 3.8B·Q4_K_M·164.7 t/s·131K ctx	Q4_K_M	9.0 GB7%	164.7 t/s	131K	S97, Runs great
Mistral Large Instruct 2407122.6BChatFunctions 122.6B·Q4_K_M·6.9 t/s	Q4_K_M	86.4 GB69%	6.9 t/s	—	C44, Usable
Qwen2.5 7B Instruct 1M7.6BChatFunctions 7.6B·Q4_K_M·112.2 t/s·1010K ctx	Q4_K_M	5.8 GB5%	112.2 t/s	1010K	S94, Runs great
Midnight Miqu 70B V1.569.0BChat 69.0B·Q4_K_M·13.2 t/s·33K ctx	Q4_K_M	47.1 GB38%	13.2 t/s	33K	B52, Decent
GLM 4 9B 04149.4BChat 9.4B·Q4_K_M·93.0 t/s·33K ctx	Q4_K_M	6.6 GB5%	93.0 t/s	33K	S90, Runs great
Meta Llama 3 8B Instruct8.0BChat 8.0B·Q4_K_M·105.7 t/s	Q4_K_M	7.9 GB6%	105.7 t/s	—	S92, Runs great
Deepseek Coder 33B Instruct33.3BChatCode 33.3B·Q4_K_M·26.9 t/s·16K ctx	Q4_K_M	24.5 GB20%	26.9 t/s	16K	A68, Runs well
Glm 4 9B Chat9.4BChatFunctions 9.4B·Q4_K_M·90.3 t/s·131K ctx	Q4_K_M	8.9 GB7%	90.3 t/s	131K	S90, Runs great
Llama 2 7B Chat HF6.7BChat 6.7B·Q4_K_M·125.8 t/s	Q4_K_M	6.9 GB6%	125.8 t/s	—	S95, Runs great
OLMo 2 0425 1B1.5BChat 1.5B·Q4_K_M·383.6 t/s·4K ctx	Q4_K_M	1.7 GB1%	383.6 t/s	4K	S100, Runs great
Wildguard7.2BChat 7.2B·Q4_K_M·117.2 t/s	Q4_K_M	7.3 GB6%	117.2 t/s	—	S94, Runs great
StableBeluga270BChat 70B·Q4_K_M·13.0 t/s·4K ctx	Q4_K_M	43.6 GB35%	13.0 t/s	4K	B52, Decent
Devstral Small 250723.6BChat 23.6B·Q4_K_M·37.7 t/s·131K ctx	Q4_K_M	17.8 GB14%	37.7 t/s	131K	A74, Runs well
Mellum2 12B A2.5B Instruct SFT12.1BChat 12.1B·Q4_K_M·119.8 t/s·131K ctx	Q4_K_M	7.8 GB6%	119.8 t/s	131K	S95, Runs great
Qwen2 57B A14B Instruct57.4BChat 57.4B·Q4_K_M·47.5 t/s·33K ctx	Q4_K_M	35.7 GB29%	47.5 t/s	33K	A79, Runs well

Mac Studio M2 Ultra specs

Specs

Chip: M2 Ultra
CPU cores: 24
GPU cores: 76
Unified memory: 128 GB
For models: 124.5 GB (macOS keeps 3.5 GB)
Memory bandwidth: 800 GB/s
FP16 (dense): 54.4 TFLOPS
Neural Engine: 32 TOPS
Released: 2023-06-13

Mac Studio M2 Ultra FAQ

How much memory is available for AI on the Mac Studio M2 Ultra (128 GB)?

About 124.5 GB of the 128 GB. macOS keeps roughly 3.5 GB for itself and other apps.

What is the largest model the Mac Studio M2 Ultra (128 GB) can run?

NVIDIA Nemotron 3 Super 120B A12B BF16 (123.6B) at Q4_K_M uses 74.7 GB, 60% of the available memory, and runs at about 31.6 tok/s.

Can the Mac Studio M2 Ultra (128 GB) run Qwen3.8 27B?

Yes. Qwen3.8 27B fits at Q4_K_M, using 17.9 GB of the 124.5 GB available, and it is the best model that runs well on this machine.

How long does the Mac Studio M2 Ultra (128 GB) take to read a long prompt?

On Qwen3.8 27B, about 21 s for a 4K-token prompt and 2.8 min for a 32K-token document before the first word appears.

Is renting a GPU cheaper than buying the Mac Studio M2 Ultra?

An A100 (80 GB) rents from $1.10/hr.

48 GB GPUs All Macs Desktops & mini PCs Mac Studio M3 Ultra Mac Studio M1 Ultra Mac Studio M5 Max MacBook Pro 14" M5 Max MacBook Pro 16" M5 Max Best coding LLMs All hardware