1× RTX PRO 6000
207tok/sMixed
1 request, mixed prompt set, with FR-Spec + NEXTN (per recipe) Source
Intelligence40with thinking
- Custom kernel required
- License unclear
Local AI · Creator
jpezzulli
4 published recipes.
Recipes
1× RTX PRO 6000
109tok/sPeak
1 request, best value reported by source Source
Intelligence34with thinkingno thinking 20
1× RTX PRO 6000
89.8tok/sPeak
1 request, best value reported by source Source
Intelligence34with thinkingOriginal model's value. Heavily compressed (~2 bit), intelligence is likely much lower.
- ≤ 3 bit: quantization may cost quality
1× RTX PRO 6000
85.6tok/sPeak
1 request, best value reported by source, context 500000 Source
Intelligence34with thinkingOriginal model's value. Heavily compressed (~2 bit), intelligence is likely much lower.
- ≤ 3 bit: quantization may cost quality