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

#ModelScore
1GPT-5.44.72
2GLM-5.14.72
3Claude Sonnet 4.54.56
4GPT-5.24.55
5Kimi K2.54.17
6DeepSeek V4 Flash4.12
7Qwen 3.5 397B A17B4.08
8MiniMax-M2.53.95
9DeepSeek V3.23.78
10Qwen 3.5 122B A10B3.71
11Qwen 3.5 35B A3B3.48

Interactive version: theaggregate.ai/benchmark?slug=a3s-bench-baseline-operators-general-security-score · How It Works · Data refreshed daily, snapshot 2026-09-25.