readygpu

Mac Studio M4 Max 64GB vs NVIDIA GeForce RTX 4090

For running models locally, priced in the US. At 64k context the first runs 57 models fully in memory, the second 40.

In short

The Mac Studio M4 Max 64GB has more memory for models: 64 GB against 24 GB. At 64k context the Mac Studio M4 Max 64GB runs 57 of the models we track fully in memory and the NVIDIA GeForce RTX 4090 40; only the Mac Studio M4 Max 64GB holds Qwen 3.5 122B-A10B. On the models both run fully in memory, the NVIDIA GeForce RTX 4090 is about 1.8× as fast as the Mac Studio M4 Max 64GB (estimated, median across models). In the US the Mac Studio M4 Max 64GB costs $4,299 at Amazon and the NVIDIA GeForce RTX 4090 $3,280 at Amazon: the NVIDIA GeForce RTX 4090 is $1,019 cheaper.

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

Mac Studio M4 Max 64GB

$4,299

Amazon

Add to build
NVIDIA GeForce RTX 4090

$3,280

Amazon

Add to build

Specifications

SpecMac Studio M4 Max 64GBNVIDIA GeForce RTX 4090
ChipApple M4 Max (40-core GPU)–
Memory bus512-bit384-bit
Memory bandwidth546 GB/s1,008 GB/s
FP16 compute16.16 TFLOPS165 TFLOPS
SoftwareMetalCUDA
Unified memory64 GB–
GPU can use by default48 GB–
GPU can use, tuned56 GB–
Operating systemmacos–
ProcessorM4 Max 16-core–
Storage1000 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 M4 Max 64GBNVIDIA GeForce RTX 4090
Qwen 3.8 27B11tok/sQ8_029tok/sQ4_K_M
Qwen 3.5 122B-A10B42tok/sUD-Q2_K_XLoffload
Qwen 3.6 27B11tok/sQ8_026tok/sQ5_K_M
Qwen3 Coder Next53tok/sQ4_K_Moffload
Gemma 4 31B9tok/sQ8_0offload
Mistral Small 4 119B55tok/sUD-Q2_K_XLoffload
Qwen 3.6 35B-A3B48tok/sQ8_0115tok/sQ4_K_M
Qwen 3.5 27B11tok/sQ8_030tok/sQ4_K_M
Gemma 4 26B-A4B37tok/sQ8_078tok/sUD-Q5_K_M
Qwen3 Next 80B-A3B61tok/sQ4_K_Moffload
Qwen 3.5 35B-A3B48tok/sQ8_0116tok/sQ4_K_M
GLM 4.7 Flash24tok/sQ8_057tok/sQ4_K_M
Devstral Small 2 24B10tok/sQ8_0offload
Nemotron 3.5 Lightning 30B-A3B56tok/sUD-Q6_K_XLoffload
Kimi Linear 48B-A3B70tok/sQ6_Koffload
gpt-oss 20B53tok/sMXFP493tok/sMXFP4
Qwen3 Coder 30B-A3B22tok/sQ8_0offload
Qwen3 32B8tok/sQ6_Koffload
Gemma 4 12B25tok/sQ8_044tok/sQ8_0
Qwen 3.5 9B30tok/sQ8_053tok/sQ8_0
Gemma 3 27B10tok/sQ8_028tok/sQ4_K_M
Mistral Small 3.2 24B10tok/sQ8_0offload
DeepSeek R1 Distill Qwen 32B8tok/sQ6_Koffload
Ministral 3 14B14tok/sQ8_029tok/sQ6_K
Qwen3 14B13tok/sQ8_027tok/sQ6_K
DeepSeek R1 0528 Qwen3 8B19tok/sQ8_034tok/sQ8_0
Qwen3 8B19tok/sQ8_034tok/sQ8_0
Phi-4 14B12tok/sQ8_026tok/sQ5_K_M
Qwen 3.5 4B53tok/sQ8_094tok/sQ8_0
Ministral 3 8B20tok/sQ8_034tok/sQ8_0
Granite 4.2 8B18tok/sQ8_031tok/sQ8_0
Gemma 4 E4B35tok/sQ8_062tok/sQ8_0
LFM2.5 8B-A1B50tok/sQ8_088tok/sQ8_0
Llama 3.1 8B23tok/sQ6_K41tok/sQ6_K
Qwen2.5 0.5B266tok/sQ8_0468tok/sQ8_0
Qwen3.5 0.8B217tok/sQ8_0382tok/sQ8_0
Qwen3.5 2B157tok/sQ5_K_M277tok/sQ5_K_M
DeepSeek R1 Distill Qwen 1.5B94tok/sQ8_0166tok/sQ8_0
MiniCPM5 2B65tok/sQ8_0114tok/sQ8_0
Mellum2.1 12B A2.5B Thinking58tok/sQ8_0103tok/sQ8_0
LFM2.5 2.6B51tok/sQ8_091tok/sQ8_0
LFM2.5 VL 3B51tok/sQ8_091tok/sQ8_0
Kimi VL A3B40tok/smlx-8bit71tok/smlx-8bit
granite vision 4.1 4b40tok/sQ8_070tok/sQ8_0
granite 4.2 3b38tok/sQ8_068tok/sQ8_0
DeepSeek R1 Distill Qwen 7B30tok/sQ8_053tok/sQ8_0
LFM2.5 24B A2B28tok/sQ8_054tok/sQ5_K_M
GLM 4.6V Flash28tok/sQ8_049tok/sQ8_0
GLM 4.1V 9B Thinking25tok/smlx-8bit43tok/smlx-8bit
DeepSeek R1 Distill Llama 8B21tok/sQ8_037tok/sQ8_0
Ministral 3 8B Reasoning 251220tok/sQ8_034tok/sQ8_0
DeepSeek R1 Distill Qwen 14B12tok/sQ8_027tok/sQ5_K_M
Devstral Small 250710tok/sQ8_0offload
Magistral Small 250610tok/sQ8_0offload
Magistral Small 250710tok/sQ8_0offload
Magistral Small 250910tok/sQ8_0offload
granite 4.2 30b7tok/sQ8_0offload
DeepSeek V4 Flashwon't runwon't run
GLM 5.3 Flashwon't runwon't run
Qwen 3.8 Flash Nextwon't runwon't run
MiniMax M2.7won't runwon't run
Nemotron 3 Super 120B-A12Bwon't runwon't run
gpt-oss 120Bwon't runwon't run
GLM 4.5 Airwon't runwon't run
Llama 3.3 70Bwon't runwon't run
DeepSeek R1won't runwon't run
DeepSeek R1 0528won't runwon't run
DeepSeek R1 Distill Llama 70Bwon't runwon't run
DeepSeek V3 0324won't runwon't run
DeepSeek V3.1won't runwon't run
DeepSeek V3.2won't runwon't run
DeepSeek V4 Flash 0731won't runwon't run
DeepSeek V4 Flash Vision Expwon't runwon't run
Devstral 2 123B Instruct 2512won't runwon't run
GLM 4.5Vwon't runwon't run
GLM 4.6Vwon't runwon't run
GLM 4.7won't runwon't run
GLM 5.1won't runwon't run
GLM 5.2won't runwon't run
GLM 5.3won't runwon't run
Kimi K2won't runwon't run
Kimi K2 Instruct 0905won't runwon't run
Kimi K2 Thinkingwon't runwon't run
Kimi K2.5won't runwon't run
MiniMax M2won't runwon't run
MiniMax M2.1won't runwon't run
MiniMax M2.5won't runwon't run
MiniMax M3won't runwon't run
Mistral Medium 3.5 128Bwon't runwon't run

Questions

Which is better for local AI, the Mac Studio M4 Max 64GB or the NVIDIA GeForce RTX 4090?
Choose the Mac Studio M4 Max 64GB 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 M4 Max 64GB or the NVIDIA GeForce RTX 4090?
At 64k context the Mac Studio M4 Max 64GB runs 57 of the models we track fully in memory and the NVIDIA GeForce RTX 4090 40; only the Mac Studio M4 Max 64GB holds Qwen 3.5 122B-A10B.
Which is faster for local AI, the Mac Studio M4 Max 64GB or the NVIDIA GeForce RTX 4090?
On the models both run fully in memory, the NVIDIA GeForce RTX 4090 is about 1.8× as fast as the Mac Studio M4 Max 64GB (estimated, median across models).
What is the price difference between the Mac Studio M4 Max 64GB and the NVIDIA GeForce RTX 4090?
In the US the Mac Studio M4 Max 64GB costs $4,299 at Amazon and the NVIDIA GeForce RTX 4090 $3,280 at Amazon: the NVIDIA GeForce RTX 4090 is $1,019 cheaper.

More comparisons