Mac Studio M3 Ultra 96GB vs NVIDIA GeForce RTX 5090
For running models locally, priced in the US. At 64k context the first runs 38 models fully in memory, the second 27.
In short
The Mac Studio M3 Ultra 96GB has more memory for models: 96 GB against 32 GB. At 64k context the Mac Studio M3 Ultra 96GB runs 38 of the models we track fully in memory and the NVIDIA GeForce RTX 5090 27; only the Mac Studio M3 Ultra 96GB holds Qwen 3.5 122B-A10B. On the models both run fully in memory, the NVIDIA GeForce RTX 5090 is about 2.1× as fast as the Mac Studio M3 Ultra 96GB (estimated, median across models). In the US the Mac Studio M3 Ultra 96GB costs $4,399 at eBay and the NVIDIA GeForce RTX 5090 $4,921 at eBay: the Mac Studio M3 Ultra 96GB is $522 cheaper.
Updated . Speeds are estimates; how they are worked out.
Specifications
| Spec | Mac Studio M3 Ultra 96GB | NVIDIA GeForce RTX 5090 |
|---|---|---|
| Chip | Apple M3 Ultra (80-core GPU) | – |
| Memory bus | 1024-bit | 512-bit |
| Memory bandwidth | 819 GB/s | 1,792 GB/s |
| FP16 compute | 28 TFLOPS | 209 TFLOPS |
| Software | Metal | CUDA |
| Unified memory | 96 GB | – |
| GPU can use by default | 72 GB | – |
| GPU can use, tuned | 88 GB | – |
| Operating system | macos | – |
| Processor | M3 Ultra 32-core | – |
| Storage | 1000 GB | – |
| GPU | – | GeForce RTX 5090 |
| Memory | – | 32 GB GDDR7 |
| Board power | – | 575 W |
| PCIe generation | – | Gen 5 |
| PCIe lanes | – | x16 |
What each runs
Estimated decode speed with the most precise quant that fits each one est.
| Model | Mac Studio M3 Ultra 96GB | NVIDIA GeForce RTX 5090 |
|---|---|---|
| Qwen 3.8 27B | 16tok/sQ8_0 | 42tok/sUD-Q6_K |
| Qwen 3.5 122B-A10B | 51tok/sQ3_K_M | offload |
| Qwen 3.6 27B | 16tok/sQ8_0 | 41tok/sQ6_K |
| Qwen3 Coder Next | 66tok/sQ6_K | offload |
| gpt-oss 120B | 58tok/sMXFP4 | won't run |
| Nemotron 3 Super 120B-A12B | 48tok/sUD-Q3_K_M | offload |
| Gemma 4 31B | 14tok/sQ8_0 | 35tok/sQ6_K |
| Mistral Small 4 119B | 54tok/sUD-Q4_K_M | offload |
| Qwen 3.6 35B-A3B | 71tok/sQ8_0 | 173tok/sUD-Q6_K |
| Qwen 3.5 27B | 16tok/sQ8_0 | 41tok/sQ6_K |
| Gemma 4 26B-A4B | 56tok/sQ8_0 | 117tok/sQ8_0 |
| GLM 4.5 Air | 15tok/sQ3_K_M | offload |
| Qwen3 Next 80B-A3B | 78tok/sQ6_K | offload |
| Qwen 3.5 35B-A3B | 71tok/sQ8_0 | 175tok/sQ6_K |
| GLM 4.7 Flash | 36tok/sQ8_0 | 87tok/sQ6_K |
| Devstral Small 2 24B | 15tok/sQ8_0 | 37tok/sQ6_K |
| Nemotron 3.5 Lightning 30B-A3B | 85tok/sUD-Q6_K_XL | offload |
| Kimi Linear 48B-A3B | 85tok/sQ8_0 | 277tok/sQ4_K_M |
| gpt-oss 20B | 79tok/sMXFP4 | 165tok/sMXFP4 |
| Qwen3 Coder 30B-A3B | 33tok/sQ8_0 | 75tok/sQ6_K |
| Qwen3 32B | 10tok/sQ8_0 | offload |
| Gemma 4 12B | 38tok/sQ8_0 | 79tok/sQ8_0 |
| Qwen 3.5 9B | 45tok/sQ8_0 | 95tok/sQ8_0 |
| Gemma 3 27B | 15tok/sQ8_0 | 40tok/sQ6_K |
| Mistral Small 3.2 24B | 15tok/sQ8_0 | 37tok/sQ6_K |
| DeepSeek R1 Distill Qwen 32B | 10tok/sQ8_0 | offload |
| Llama 3.3 70B | 8tok/sQ5_K_M | won't run |
| Ministral 3 14B | 21tok/sQ8_0 | 44tok/sQ8_0 |
| Qwen3 14B | 20tok/sQ8_0 | 42tok/sQ8_0 |
| DeepSeek R1 0528 Qwen3 8B | 29tok/sQ8_0 | 61tok/sQ8_0 |
| Qwen3 8B | 29tok/sQ8_0 | 61tok/sQ8_0 |
| Phi-4 14B | 18tok/sQ8_0 | 38tok/sQ8_0 |
| Qwen 3.5 4B | 80tok/sQ8_0 | 166tok/sQ8_0 |
| Ministral 3 8B | 29tok/sQ8_0 | 61tok/sQ8_0 |
| Granite 4.2 8B | 27tok/sQ8_0 | 55tok/sQ8_0 |
| Gemma 4 E4B | 53tok/sQ8_0 | 110tok/sQ8_0 |
| LFM2.5 8B-A1B | 75tok/sQ8_0 | 156tok/sQ8_0 |
| Llama 3.1 8B | 35tok/sQ6_K | 73tok/sQ6_K |
| DeepSeek V4 Flash | won't run | won't run |
| GLM 5.3 Flash | won't run | won't run |
| Qwen 3.8 Flash Next | won't run | won't run |
| MiniMax M2.7 | won't run | won't run |
Questions
- Which is better for local AI, the Mac Studio M3 Ultra 96GB or the NVIDIA GeForce RTX 5090?
- Choose the Mac Studio M3 Ultra 96GB to run bigger models. Choose the NVIDIA GeForce RTX 5090 for speed. Choose the Mac Studio M3 Ultra 96GB to spend less.
- Which can run bigger models, the Mac Studio M3 Ultra 96GB or the NVIDIA GeForce RTX 5090?
- At 64k context the Mac Studio M3 Ultra 96GB runs 38 of the models we track fully in memory and the NVIDIA GeForce RTX 5090 27; only the Mac Studio M3 Ultra 96GB holds Qwen 3.5 122B-A10B.
- Which is faster for local AI, the Mac Studio M3 Ultra 96GB or the NVIDIA GeForce RTX 5090?
- On the models both run fully in memory, the NVIDIA GeForce RTX 5090 is about 2.1× as fast as the Mac Studio M3 Ultra 96GB (estimated, median across models).
- What is the price difference between the Mac Studio M3 Ultra 96GB and the NVIDIA GeForce RTX 5090?
- In the US the Mac Studio M3 Ultra 96GB costs $4,399 at eBay and the NVIDIA GeForce RTX 5090 $4,921 at eBay: the Mac Studio M3 Ultra 96GB is $522 cheaper.
More comparisons
- ASUS Ascent GX10 128GB vs NVIDIA GeForce RTX 5090
- Framework Desktop Ryzen AI Max+ 395 128GB vs NVIDIA GeForce RTX 5090
- GMKtec EVO-X2 Ryzen AI Max+ 395 128GB vs Mac Studio M3 Ultra 96GB
- GMKtec EVO-X2 Ryzen AI Max+ 395 128GB vs NVIDIA GeForce RTX 5090
- Mac Studio M3 Ultra 256GB vs Mac Studio M3 Ultra 96GB
- Mac Studio M3 Ultra 512GB vs Mac Studio M3 Ultra 96GB
- Mac Studio M3 Ultra 96GB vs Mac Studio M4 Max 128GB
- Mac Studio M3 Ultra 96GB vs Mac Studio M4 Max 64GB
- Mac Studio M3 Ultra 96GB vs NVIDIA RTX PRO 6000 Blackwell
- Mac Studio M4 Max 128GB vs NVIDIA GeForce RTX 5090