← All models
Creator
NVIDIA
4 models in the catalogue
Every model, most used across the global market first.
4 models
NVIDIA
| The cost of a typical message: 2,000 input and 500 output tokens. Reasoning models count their thinking as output too, so a message costs more. | The most text the model reads at once | The Artificial Analysis intelligence index, 0–100 | How fast the answer is written, in tokens per second | Place by usage across the global model market, last 7 days | ||||
|---|---|---|---|---|---|---|---|---|
| 1.64M | 6.54M | 6,542 | 262K | 11.8 | 68From public market statistics for this model | 99.97%Healthy | #16Down 9 places on last week | |
| 191K | 545K | 654 | 262K | 5.5 | 119From public market statistics for this model | 94.7%Healthy | #30Down 2 places on last week | |
| 164K | 654K | 654 | 262K | — | 160From public market statistics for this model | 67.7%Unavailable | Unranked | |
| 818K | 6.54M | 4,907 | 1M | — | 75From public market statistics for this model | 99.9%Healthy | Unranked |
NVIDIA: Nemotron 3 UltraNVIDIA · 262K tokens6,542IRR a message
99.97%HealthyIntelligence 11.8#16Down 9 places on last week
NVIDIA: Nemotron 3.5 LightningNVIDIA · 262K tokens654IRR a message
94.7%HealthyIntelligence 5.5#30Down 2 places on last week
1–4 of 4