CESBench - Multiple Choice: leaderboard
The 209 multiple-choice items of CESBench (Zhou et al., 2026) on cryptographic engineering security for IoT devices, with distractors that encode documented practitioner misconceptions. Accuracy in percent, zero-shot at temperature 0; higher is better.
Metric: Accuracy (%; the 209 four-option multiple-choice items whose distractors encode practitioner misconceptions; 380 expert-written items on cryptographic engineering security for IoT devices; zero-shot, temperature 0, no tools, provider default reasoning mode; judgment and scenario answers scored by a Qwen3.5-397B judge from outside the evaluated set). Source: arxiv.org. Saturation forecast: Estimated already saturated. 11 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GLM-5.2 | 98.6 | #137 |
| 2 | Gemini 3.7 Flash | 98.6 | #32 |
| 3 | GPT-5.6 Luna | 97.6 | #117 |
| 4 | DeepSeek V4 Pro | 97.1 | #181 |
| 5 | Kimi K2.6 | 97.1 | #173 |
| 6 | MiniMax-M2.5 | 95.2 | #590 |
| 7 | Llama 4 Maverick | 93.3 | #1105 |
| 8 | DeepSeek V3 | 91.9 | #705 |
| 9 | GLM-4 32B (0414) | 88.5 | #1407 |
| 10 | Hunyuan A13B-Instruct | 85.6 | #2021 |
| 11 | Ling-flash-2.0 | 83.3 | #1352 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=cesbench-multiple-choice · How It Works · Data refreshed daily, snapshot 2026-11-12.