BioSecBench-Refusal: leaderboard
Metric: Red-Team total refusal (%). Source: benchmarklist.com. 16 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 4.7 | 62.3 |
| 2 | Claude Opus 4.8 | 53.6 |
| 3 | Claude Sonnet 4.6 | 50 |
| 4 | Gemini 3.5 Flash | 46.5 |
| 5 | Claude Opus 4.6 | 41.7 |
| 6 | GPT-5.4 | 35.1 |
| 7 | Gemini 3.1 Pro (Preview) | 28.1 |
| 8 | GPT-5.5 | 23.2 |
| 9 | Grok 4.20 Beta (0309) (Reasoning) | 1.4 |
Interactive version: theaggregate.ai/benchmark?slug=biosecbench-refusal · How It Works · Data refreshed daily, snapshot 2026-09-19.