"The architecture is incredibly fast."
"5x faster."
"30% cheaper."
Here's the new AI.
9:19 PM · May 31, 2026 · 4.3K Views
Artificial Analysis clocks the model at 280 tokens/sec.
"The architecture is incredibly fast."
"5x faster."
"30% cheaper."
Here's the new AI.
Artificial Analysis clocks the model at 280 tokens/sec.
"The architecture is incredibly fast."
"5x faster."
"30% cheaper."
Here's the new AI.
Many users praised NVIDIA and Jensen for the Nemotron 3 Ultra model's 5x faster speed and 30% lower costs, while some dismissed the claims as typical unconvincing hype.