NVIDIA
Comparing 3 models from NVIDIA.
| Model | Total ▼ | Released | Context | CoT | Tooling | Creative Writing | Language | Utility | Reasoning | Text Editing | Rule Following | Hallucination |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Nemotron 3 Super | 84.56% | Mar 11, 26 | 262k | – | 93.49% | 69.75% | 81.41% | 95.29% | 93.11% | 86.34% | 57.43% | 99.69% |
| Nemotron 3 Nano | 77.73% | Dec 14, 25 | 262k | – | 83.98% | 65.87% | 87.63% | 86.00% | 89.91% | 75.81% | 43.47% | 89.17% |
| Llama 3.1 Nemotron 70B | 74.70% | Oct 15, 24 | 128k | – | 95.74% | 71.71% | 46.80% | 88.31% | 82.19% | 87.26% | 50.62% | 74.99% |
Model Performance
Cost vs Performance
Compares total benchmark cost against overall score for NVIDIA models. Quadrant lines are drawn at the median values.
Cost Breakdown
Total benchmark cost per model, broken down by input, reasoning, and output tokens. Toggle between USD and token views.