OI-Bench - Bonus Framing - Attack Success Rate: leaderboard

Metric: Attack Success Rate (%). Source: arxiv.org. Saturation forecast: Estimated already saturated. 12 models tracked.

Top models

#ModelScore
1GPT-53.76
2Claude Haiku 4.54.6
3Grok 4.15.99
4GPT-5 Mini6.31
5Llama 4 Scout6.4
6Llama 4 Maverick7.15
7Qwen 3 8B11.86
8Qwen 3 235B A22B13.67
9DeepSeek V3.216.96
10DeepSeek R118.5
11Gemini 2.5 Pro26.28
12Gemini 2.5 Flash Lite31.36

Interactive version: theaggregate.ai/benchmark?slug=oi-bench-bonus-framing-attack-success-rate · How It Works · Data refreshed daily, snapshot 2026-09-25.