RTX 5090 vs RTX Pro 6000 Blackwell
For running local LLMs · 92 reports across 15 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 | RTX 5090 | RTX Pro 6000 Blackwell |
|---|---|---|
| Qwen3.8up to 2400B | 220.0n=35 | 177.0n=13 |
| DeepSeek V4 Flashup to 285B | 120.0n=10 | 45.0n=4 |
| Qwen3.6up to 35B | 215.1n=6 | —n=4 |
| Museup to 30B | 253.0n=4 | 57.0n=2 |
| Gemma 4up to 31B | 578.0n=4 | — |
| Aurora1.0up to 0B | — | —n=1 |
| GLM-5.2up to 744B | —n=1 | — |
| GLM-5.3up to 321B | 24.5n=1 | — |
| Kimi K2.5 | —n=1 | — |
| Llama 3.2up to 1B | — | —n=1 |
| Mimo 2.5 | — | —n=1 |
| Qwen2.5up to 32B | —n=1 | — |
| Qwen3up to 235B | 159.0n=1 | — |
| Qwen3.5up to 122B | —n=1 | — |
| Tencent-HY3up to 295B | — | 57.4n=1 |