HypoArena - Safety Investigation: leaderboard
Metric: Rating on the 103 safety investigation cases: Bradley-Terry-Davidson rating (base 1500, open-ended) from reference-independent pairwise judgments by seed-2.0-pro, both presentation orders, with the source-derived reference hypothesis set as an anonymous competitor; baseline single-pass generation; higher is better. Source: arxiv.org. Saturation forecast: Around September 2027. 15 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Sonnet 4.6 (High) | 1721.6 |
| 2 | Claude Opus 4.6 (High) | 1668.4 |
| 3 | Kimi K2.6 | 1633.5 |
| 4 | GPT-5.4 (High) | 1630.7 |
| 5 | GLM-5.1 | 1560 |
| 6 | DeepSeek V4 Pro (High) | 1551.4 |
| 7 | DeepSeek V4 Flash (High) | 1489.3 |
| 8 | MiniMax-M2.7 | 1453.9 |
| 9 | GLM-5 | 1445.2 |
| 10 | GPT-5.4 Mini (High) | 1437.7 |
| 11 | Gemini 3.1 Pro (Preview) (High) | 1387.2 |
| 12 | MiniMax-M2.5 | 1369.7 |
| 13 | Kimi K2.5 (Thinking) | 1289.4 |
| 14 | Gemini 3 Flash (High) | 1262.1 |
Interactive version: theaggregate.ai/benchmark?slug=hypoarena-safety-investigation · How It Works · Data refreshed daily, snapshot 2026-09-29.