NuclearQAv2 - Boolean: leaderboard
Yes-or-no factual questions of NuclearQAv2 (2026), a nuclear engineering question-answering benchmark: accuracy in percent averaged over three runs. Higher is better.
Metric: Accuracy (%) on 750 yes-or-no nuclear engineering questions (factual grounding); mean of three runs; answers requested without reasoning text. Source: arxiv.org. Saturation forecast: Estimated already saturated. 9 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.2 | 80.22 | #176 |
| 2 | GPT-5.4 | 79.91 | #123 |
| 3 | GPT-OSS-120B | 79.6 | #700 |
| 4 | NVIDIA-Nemotron-3-Super-120B-A12B-FP8 | 79.24 | #469 |
| 5 | Llama 3 70B Instruct | 76 | #1676 |
| 6 | GPT-4o | 73.07 | #741 |
| 7 | Llama 3 8B Instruct | 71.73 | #4612 |
| 8 | Nova Pro | 56.98 | #1555 |
| 9 | Mistral 7B Instruct | 38.31 | #7468 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=nuclearqav2-boolean · How It Works · Data refreshed daily, snapshot 2026-11-13.