A3S-Bench - Baseline Operators - General Security Score: leaderboard
Metric: General Security Score (1-5 judge rating of the full trace; 5 = safe). Source: arxiv.org. Saturation forecast: Around May 2027. 11 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.4 | 4.72 |
| 2 | GLM-5.1 | 4.72 |
| 3 | Claude Sonnet 4.5 | 4.56 |
| 4 | GPT-5.2 | 4.55 |
| 5 | Kimi K2.5 | 4.17 |
| 6 | DeepSeek V4 Flash | 4.12 |
| 7 | Qwen 3.5 397B A17B | 4.08 |
| 8 | MiniMax-M2.5 | 3.95 |
| 9 | DeepSeek V3.2 | 3.78 |
| 10 | Qwen 3.5 122B A10B | 3.71 |
| 11 | Qwen 3.5 35B A3B | 3.48 |
Interactive version: theaggregate.ai/benchmark?slug=a3s-bench-baseline-operators-general-security-score · How It Works · Data refreshed daily, snapshot 2026-09-25.