M1 Ultra 128GB vs RTX Pro 6000 Blackwell
For running local LLMs · 29 reports across 9 models
Highest reported generation speed by model family
These maxima come from different community setups. Model size, quantization, GPU count, offloading and concurrency may differ. They do not establish which GPU is faster for the same workload. Check the comparison conditions →
| Model | M1 Ultra 128GB | RTX Pro 6000 Blackwell |
|---|---|---|
| Qwen3.8up to 2400B | 31.2n=1 | 177.0n=13 |
| DeepSeek V4 Flashup to 285B | — | 45.0n=4 |
| Qwen3.6up to 35B | — | —n=4 |
| Museup to 30B | — | 57.0n=2 |
| Aurora1.0up to 0B | — | —n=1 |
| DeepSeek V4up to 284B | 16.0n=1 | — |
| Llama 3.2up to 1B | — | —n=1 |
| Mimo 2.5 | — | —n=1 |
| Tencent-HY3up to 295B | — | 57.4n=1 |