Mac Pro M2 Ultra 192GB
Apple · M2 Ultra · 192GB Unified Memory · Can run 25 models
| Manufacturer | Apple |
| Unified Memory | 192 GB |
| Chip | M2 Ultra |
| CPU Cores | 24 |
| GPU Cores | 76 |
| Neural Engine Cores | 32 |
| Memory Bandwidth | 800 GB/s |
| MSRP | $6,999 |
| Released | Jun 13, 2023 |
AI Notes
The Mac Pro M2 Ultra 192GB offers massive memory capacity for large AI models in a tower form factor. With 192GB of unified memory and 800 GB/s bandwidth, it can run full-precision 70B models and load 100B+ parameter models with quantization. Its expandable tower design supports additional storage and connectivity for professional AI development workflows.
Compatible Models
| Model | Parameters | Best Quant | VRAM Used | Fit |
|---|---|---|---|---|
| Llama 3.2 1B | 1B | Q8_0 | 3 GB | Runs |
| Gemma 2 2B | 2B | Q8_0 | 4 GB | Runs |
| Llama 3.2 3B | 3B | Q8_0 | 5 GB | Runs |
| Phi-3 Mini 3.8B | 3.8B | Q8_0 | 5.8 GB | Runs |
| DeepSeek R1 7B | 7B | Q8_0 | 9 GB | Runs |
| Mistral 7B | 7B | Q8_0 | 9 GB | Runs |
| Qwen 2.5 7B | 7B | Q8_0 | 9 GB | Runs |
| Qwen 2.5 Coder 7B | 7B | Q8_0 | 9 GB | Runs |
| Llama 3.1 8B | 8B | Q8_0 | 10 GB | Runs |
| Gemma 2 9B | 9B | Q8_0 | 11 GB | Runs |
| DeepSeek R1 14B | 14B | Q4_K_M | 9.9 GB | Runs |
| Phi-4 14B | 14B | Q4_K_M | 9.9 GB | Runs |
| Qwen 2.5 14B | 14B | Q4_K_M | 9.9 GB | Runs |
| StarCoder2 15B | 15B | Q8_0 | 17 GB | Runs |
| Codestral 22B | 22B | Q4_K_M | 14.7 GB | Runs |
| Gemma 2 27B | 27B | Q4_K_M | 17.7 GB | Runs |
| DeepSeek R1 32B | 32B | Q4_K_M | 20.7 GB | Runs |
| Qwen 2.5 32B | 32B | Q4_K_M | 20.7 GB | Runs |
| Command R 35B | 35B | Q4_K_M | 22.5 GB | Runs |
| Mixtral 8x7B | 47B | Q4_K_M | 29.7 GB | Runs |
| DeepSeek R1 70B | 70B | Q4_K_M | 43.5 GB | Runs |
| Llama 3.1 70B | 70B | Q4_K_M | 43.5 GB | Runs |
| Llama 3.3 70B | 70B | Q4_K_M | 43.5 GB | Runs |
| Qwen 2.5 72B | 72B | Q4_K_M | 44.7 GB | Runs |
| Llama 3.1 405B | 405B | Q4_K_M | 244.5 GB | CPU Offload |