Mac Studio M3 Ultra 512GB vs Mac Studio M3 Ultra 96GB
For running models locally, priced in the US. At 128k context the first runs 42 models fully in memory, the second 37.
In short
The Mac Studio M3 Ultra 512GB has more memory for models: 512 GB against 96 GB. At 128k context the Mac Studio M3 Ultra 512GB runs 42 of the models we track fully in memory and the Mac Studio M3 Ultra 96GB 37; only the Mac Studio M3 Ultra 512GB holds DeepSeek V4 Flash. On the models both run fully in memory, the Mac Studio M3 Ultra 512GB and the Mac Studio M3 Ultra 96GB are about as fast as each other (estimated). In the US the Mac Studio M3 Ultra 512GB costs $23,000 at eBay and the Mac Studio M3 Ultra 96GB $4,399 at eBay: the Mac Studio M3 Ultra 96GB is $18,601 cheaper.
Updated . Speeds are estimates; how they are worked out.
Specifications
| Spec | Mac Studio M3 Ultra 512GB | Mac Studio M3 Ultra 96GB |
|---|---|---|
| Chip | Apple M3 Ultra (80-core GPU) | Apple M3 Ultra (80-core GPU) |
| Memory bus | 1024-bit | 1024-bit |
| Memory bandwidth | 819 GB/s | 819 GB/s |
| FP16 compute | 28 TFLOPS | 28 TFLOPS |
| Software | Metal | Metal |
| Unified memory | 512 GB | 96 GB |
| GPU can use by default | 384 GB | 72 GB |
| GPU can use, tuned | 504 GB | 88 GB |
| Operating system | macos | macos |
| Processor | M3 Ultra 32-core | M3 Ultra 32-core |
| Storage | 2000 GB | 1000 GB |
What each runs
Estimated decode speed with the most precise quant that fits each one est.
| Model | Mac Studio M3 Ultra 512GB | Mac Studio M3 Ultra 96GB |
|---|---|---|
| DeepSeek V4 Flash | 14tok/sUD-Q4_K_XL | won't run |
| GLM 5.3 Flash | 11tok/sQ8_0 | won't run |
| Qwen 3.8 Flash Next | 5tok/sQ8_0 | won't run |
| Qwen 3.8 27B | 14tok/sQ8_0 | 14tok/sQ8_0 |
| MiniMax M2.7 | 7tok/sQ8_0 | won't run |
| Qwen 3.5 122B-A10B | 23tok/sQ8_0 | 41tok/sQ3_K_M |
| Qwen 3.6 27B | 14tok/sQ8_0 | 14tok/sQ8_0 |
| Qwen3 Coder Next | 43tok/sQ8_0 | 49tok/sQ6_K |
| gpt-oss 120B | 40tok/sMXFP4 | 40tok/sMXFP4 |
| Nemotron 3 Super 120B-A12B | 23tok/sQ8_0 | 44tok/sUD-Q3_K_M |
| Gemma 4 31B | 12tok/sQ8_0 | 12tok/sQ8_0 |
| Mistral Small 4 119B | 30tok/sQ8_0 | 51tok/sUD-Q3_K_M |
| Qwen 3.6 35B-A3B | 55tok/sQ8_0 | 55tok/sQ8_0 |
| Qwen 3.5 27B | 14tok/sQ8_0 | 14tok/sQ8_0 |
| Gemma 4 26B-A4B | 45tok/sQ8_0 | 45tok/sQ8_0 |
| GLM 4.5 Air | 8tok/sQ8_0 | 10tok/sQ2_K |
| Qwen 3.5 35B-A3B | 55tok/sQ8_0 | 55tok/sQ8_0 |
| Qwen3 Next 80B-A3B | 50tok/sQ8_0 | 56tok/sQ6_K |
| GLM 4.7 Flash | 26tok/sQ8_0 | 26tok/sQ8_0 |
| Devstral Small 2 24B | 11tok/sQ8_0 | 11tok/sQ8_0 |
| Nemotron 3.5 Lightning 30B-A3B | 76tok/sUD-Q6_K_XL | 76tok/sUD-Q6_K_XL |
| Kimi Linear 48B-A3B | 75tok/sQ8_0 | 75tok/sQ8_0 |
| gpt-oss 20B | 57tok/sMXFP4 | 57tok/sMXFP4 |
| Qwen3 Coder 30B-A3B | 20tok/sQ8_0 | 20tok/sQ8_0 |
| Qwen3 32B | 8tok/sQ8_0 | 8tok/sQ8_0 |
| Gemma 4 12B | 35tok/sQ8_0 | 35tok/sQ8_0 |
| Qwen 3.5 9B | 38tok/sQ8_0 | 38tok/sQ8_0 |
| Gemma 3 27B | 13tok/sQ8_0 | 13tok/sQ8_0 |
| Mistral Small 3.2 24B | 11tok/sQ8_0 | 11tok/sQ8_0 |
| DeepSeek R1 Distill Qwen 32B | 8tok/sQ8_0 | 8tok/sQ8_0 |
| Llama 3.3 70B | 5tok/sQ8_0 | won't run |
| Ministral 3 14B | 15tok/sQ8_0 | 15tok/sQ8_0 |
| Qwen3 14B | 14tok/sQ8_0 | 14tok/sQ8_0 |
| DeepSeek R1 0528 Qwen3 8B | 19tok/sQ8_0 | 19tok/sQ8_0 |
| Qwen3 8B | 19tok/sQ8_0 | 19tok/sQ8_0 |
| Phi-4 14B | 13tok/sQ8_0 | 13tok/sQ8_0 |
| Qwen 3.5 4B | 60tok/sQ8_0 | 60tok/sQ8_0 |
| Ministral 3 8B | 20tok/sQ8_0 | 20tok/sQ8_0 |
| Granite 4.2 8B | 17tok/sQ8_0 | 17tok/sQ8_0 |
| Gemma 4 E4B | 44tok/sQ8_0 | 44tok/sQ8_0 |
| LFM2.5 8B-A1B | 43tok/sQ8_0 | 43tok/sQ8_0 |
| Llama 3.1 8B | 22tok/sQ6_K | 22tok/sQ6_K |
Questions
- Which is better for local AI, the Mac Studio M3 Ultra 512GB or the Mac Studio M3 Ultra 96GB?
- Choose the Mac Studio M3 Ultra 512GB to run bigger models. Choose the Mac Studio M3 Ultra 96GB to spend less.
- Which can run bigger models, the Mac Studio M3 Ultra 512GB or the Mac Studio M3 Ultra 96GB?
- At 128k context the Mac Studio M3 Ultra 512GB runs 42 of the models we track fully in memory and the Mac Studio M3 Ultra 96GB 37; only the Mac Studio M3 Ultra 512GB holds DeepSeek V4 Flash.
- Which is faster for local AI, the Mac Studio M3 Ultra 512GB or the Mac Studio M3 Ultra 96GB?
- On the models both run fully in memory, the Mac Studio M3 Ultra 512GB and the Mac Studio M3 Ultra 96GB are about as fast as each other (estimated).
- What is the price difference between the Mac Studio M3 Ultra 512GB and the Mac Studio M3 Ultra 96GB?
- In the US the Mac Studio M3 Ultra 512GB costs $23,000 at eBay and the Mac Studio M3 Ultra 96GB $4,399 at eBay: the Mac Studio M3 Ultra 96GB is $18,601 cheaper.