Mac Studio M3 Ultra 256GB vs NVIDIA GeForce RTX 4090

For running models locally, priced in the UK. At 8k context the first runs 42 models fully in memory, the second 28.

In short

The Mac Studio M3 Ultra 256GB has more memory for models: 256 GB against 24 GB. At 8k context the Mac Studio M3 Ultra 256GB runs 42 of the models we track fully in memory and the NVIDIA GeForce RTX 4090 28; only the Mac Studio M3 Ultra 256GB holds DeepSeek V4 Flash. On the models both run fully in memory, the NVIDIA GeForce RTX 4090 is about 1.3× as fast as the Mac Studio M3 Ultra 256GB (estimated, median across models). In the UK the Mac Studio M3 Ultra 256GB costs £8,499 at eBay (used) and the NVIDIA GeForce RTX 4090 £6,015 at Amazon: the NVIDIA GeForce RTX 4090 is £2,484 cheaper.

Updated . Speeds are estimates; how they are worked out.

Specifications

SpecMac Studio M3 Ultra 256GBNVIDIA GeForce RTX 4090
ChipApple M3 Ultra (80-core GPU)–
Memory bus1024-bit384-bit
Memory bandwidth819 GB/s1,008 GB/s
FP16 compute28 TFLOPS165 TFLOPS
SoftwareMetalCUDA
Unified memory256 GB–
GPU can use by default192 GB–
GPU can use, tuned248 GB–
Operating systemmacos–
ProcessorM3 Ultra 32-core–
Storage2000 GB–
GPU–GeForce RTX 4090
Memory–24 GB GDDR6X
Board power–450 W
PCIe generation–Gen 4
PCIe lanes–x16

What each runs

Estimated decode speed with the most precise quant that fits each one est.

ModelMac Studio M3 Ultra 256GBNVIDIA GeForce RTX 4090
DeepSeek V4 Flash28tok/sUD-Q4_K_XLwon't run
GLM 5.3 Flash20tok/sUD-Q4_K_XLwon't run
Qwen 3.8 Flash Next5tok/sQ8_0won't run
MiniMax M2.729tok/sUD-Q6_Kwon't run
Qwen 3.8 27B18tok/sQ8_028tok/sUD-Q6_K
Qwen 3.5 122B-A10B30tok/sQ8_0offload
Qwen 3.6 27B18tok/sQ8_027tok/sQ6_K
gpt-oss 120B93tok/sMXFP4won't run
Qwen3 Coder Next73tok/sQ8_0offload
Nemotron 3 Super 120B-A12B25tok/sQ8_0won't run
Gemma 4 31B16tok/sQ8_027tok/sQ5_K_M
Mistral Small 4 119B42tok/sQ8_0offload
Qwen 3.6 35B-A3B97tok/sQ8_0181tok/sUD-Q4_K_M
Qwen 3.5 27B18tok/sQ8_027tok/sQ6_K
Gemma 4 26B-A4B70tok/sQ8_095tok/sUD-Q6_K
GLM 4.5 Air16tok/sQ8_0offload
Qwen 3.5 35B-A3B97tok/sQ8_0181tok/sQ4_K_M
Qwen3 Next 80B-A3B94tok/sQ8_0offload
GLM 4.7 Flash56tok/sQ8_094tok/sQ5_K_M
Devstral Small 2 24B20tok/sQ8_030tok/sQ6_K
Nemotron 3.5 Lightning 30B-A3B93tok/sUD-Q6_K_XLoffload
gpt-oss 20B121tok/sMXFP4142tok/sMXFP4
Kimi Linear 48B-A3B97tok/sQ8_0offload
Qwen3 Coder 30B-A3B80tok/sQ8_0128tok/sQ5_K_M
Qwen3 32B14tok/sQ8_029tok/sQ4_K_M
Gemma 4 12B41tok/sQ8_048tok/sQ8_0
Qwen 3.5 9B54tok/sQ8_064tok/sQ8_0
Mistral Small 3.2 24B20tok/sQ8_030tok/sQ6_K
Gemma 3 27B18tok/sQ8_027tok/sQ6_K
DeepSeek R1 Distill Qwen 32B14tok/sQ8_028tok/sQ4_K_M
Llama 3.3 70B7tok/sQ8_0offload
Ministral 3 14B34tok/sQ8_040tok/sQ8_0
Qwen3 14B31tok/sQ8_037tok/sQ8_0
DeepSeek R1 0528 Qwen3 8B54tok/sQ8_063tok/sQ8_0
Qwen3 8B54tok/sQ8_063tok/sQ8_0
Phi-4 14B31tok/sQ8_036tok/sQ8_0
Qwen 3.5 4B111tok/sQ8_0130tok/sQ8_0
Ministral 3 8B52tok/sQ8_062tok/sQ8_0
Granite 4.2 8B50tok/sQ8_059tok/sQ8_0
Gemma 4 E4B63tok/sQ8_074tok/sQ8_0
LFM2.5 8B-A1B218tok/sQ8_0256tok/sQ8_0
Llama 3.1 8B69tok/sQ6_K82tok/sQ6_K

Questions

Which is better for local AI, the Mac Studio M3 Ultra 256GB or the NVIDIA GeForce RTX 4090?
Choose the Mac Studio M3 Ultra 256GB to run bigger models. Choose the NVIDIA GeForce RTX 4090 for speed. Choose the NVIDIA GeForce RTX 4090 to spend less.
Which can run bigger models, the Mac Studio M3 Ultra 256GB or the NVIDIA GeForce RTX 4090?
At 8k context the Mac Studio M3 Ultra 256GB runs 42 of the models we track fully in memory and the NVIDIA GeForce RTX 4090 28; only the Mac Studio M3 Ultra 256GB holds DeepSeek V4 Flash.
Which is faster for local AI, the Mac Studio M3 Ultra 256GB or the NVIDIA GeForce RTX 4090?
On the models both run fully in memory, the NVIDIA GeForce RTX 4090 is about 1.3× as fast as the Mac Studio M3 Ultra 256GB (estimated, median across models).
What is the price difference between the Mac Studio M3 Ultra 256GB and the NVIDIA GeForce RTX 4090?
In the UK the Mac Studio M3 Ultra 256GB costs £8,499 at eBay (used) and the NVIDIA GeForce RTX 4090 £6,015 at Amazon: the NVIDIA GeForce RTX 4090 is £2,484 cheaper.

More comparisons