Quality
Highest overall scores
- 1
NVIDIA: Nemotron 3 Ultra38.3 - 2NVIDIA: Nemotron 3.5 Lightning23.6
- 3NVIDIA: Nemotron 3 Nano 30B A3B14.5
NVIDIA model benchmarks.
Compare NVIDIA models across answer quality, coding, complex tasks, speed, price, and context size.
Response speed
Fastest output in this sample
- 1NVIDIA: Nemotron 3 Nano 30B A3B135 t/s
- 2NVIDIA: Nemotron 3.5 Lightning132 t/s
- 3NVIDIA: Nemotron 3 Ultra121 t/s
Input price
Lowest price per 1M tokens
- 1NVIDIA: Nemotron 3 Nano 30B A3B$0.05/1M
- 2NVIDIA: Nemotron 3.5 Lightning$0.10/1M
- 3NVIDIA: Nemotron 3 Ultra$0.60/1M
Highlights
Quality
Overall capability · Higher is better
Speed
Output tokens per second · Higher is better
Input price
USD per 1M tokens · Lower is better
Quality benchmarks
Compare NVIDIA model capability.
Performance and price
Compare practical tradeoffs.
All NVIDIA models.
3 models in the current catalogue.
| Model | Links | |||||
|---|---|---|---|---|---|---|
| NVIDIA: Nemotron 3 Ultra | 38.3 | 121 t/s | 512.3K | $0.60 | $3.60 | |
| NVIDIA: Nemotron 3.5 Lightning | 23.6 | 132 t/s | 1M | $0.10 | $0.25 | |
| NVIDIA: Nemotron 3 Nano 30B A3B | 14.5 | 135 t/s | 262.1K | $0.05 | $0.20 |
NVIDIA technical benchmarks.
Compare named evaluation records available for NVIDIA models. Missing results remain blank rather than being estimated.
| Benchmark | Area | Nemotron 3 Ultra | Nemotron 3.5 Lightning | Nemotron 3 Nano 30B A3B | Source |
|---|---|---|---|---|---|
| Intelligence Index | overallQuality | 38.3 | 23.6 | 14.5 | OpenRouter / artificial-analysis |
| Coding Index | coding | 49.3 | 26.8 | 14.4 | OpenRouter / artificial-analysis |
| Agentic Index | agentic | 27.5 | 13.8 | 2 | OpenRouter / artificial-analysis |
| GPQA | GPQA Diamond | 86.7 | 74.3 | 75.7 | OpenRouter model benchmarks |
| Humanity's Last Exam | HLE | 28.4 | 10.6 | 11.4 | OpenRouter model benchmarks |
| IFBench | IFBench | 81.4 | — | 71.1 | OpenRouter model benchmarks |
| τ²-Bench Telecom | τ²-Bench Telecom | 83.3 | — | 40.9 | OpenRouter model benchmarks |
| AA-LCR | AA-LCR | 71 | 55.3 | 37.3 | OpenRouter model benchmarks |
| GDPval-AA | GDPval-AA | 33.2 | 16.2 | 0 | OpenRouter model benchmarks |
| CritPt | CritPt | 3.1 | 0 | 0.9 | OpenRouter model benchmarks |
| SciCode | SciCode | 39.9 | 31.6 | 29.6 | OpenRouter model benchmarks |
| Terminal-Bench Hard | Terminal-Bench Hard | 36.4 | — | 13.6 | OpenRouter model benchmarks |
| AA-Omniscience Accuracy | AA-Omniscience Accuracy | 22.6 | 14.4 | 17.3 | OpenRouter model benchmarks |
| AA-Omniscience Non-Hallucination Rate | AA-Omniscience Non-Hallucination Rate | 70.3 | 62.4 | 16.7 | OpenRouter model benchmarks |
| MMLU-Pro | Knowledge and reasoning | — | — | 78.3 | OpenEvals/leaderboard-data |