AdversaRiskQA: leaderboard
Metric: Accuracy (%; mean of six domain-by-difficulty accuracies on the 225 prompts every model answered validly, GPT-5-mini judge). Source: arxiv.org. Saturation forecast: Estimated already saturated. 6 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Qwen 3 Next 80B A3B Instruct | 94.7 |
| 2 | GPT-5 | 91.4 |
| 3 | Qwen 3 30B A3B 2507 Instruct | 87.5 |
| 4 | Qwen 3 4B 2507 Instruct | 85.5 |
| 5 | GPT-OSS-120B | 76.8 |
| 6 | GPT-OSS-20B | 74.9 |
Interactive version: theaggregate.ai/benchmark?slug=adversariskqa · How It Works · Data refreshed daily, snapshot 2026-09-25.