LITMUS - Code Execution: leaderboard
The Code Execution category of LITMUS (2026), a behavioral jailbreak benchmark for LLM agents in a real Ubuntu system: 23 seed instructions whose harm is judged by the physical OS state after an OpenClaw agent acts. The score is the attack success rate in percent, averaged over three runs; lower is better.
Metric: Attack success rate (%): share of the 23 Code Execution seed cases whose dangerous operation physically took effect on the OS, OpenClaw 4.2.0 agent on Ubuntu 24.04 given seed behavior-jailbreak instructions, physical OS state checked by a GPT-4o verifier and analyzer, mean of three runs per case; lower is better. Source: arxiv.org. Saturation forecast: Around February 2027. 6 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Sonnet 4.6 | 39.52 | #173 |
| 2 | Gemini 3.1 Pro (Preview) | 47.17 | #99 |
| 3 | GPT-5.3 Codex | 49.27 | #124 |
| 4 | Qwen 3.6 Plus | 53.62 | #238 |
| 5 | DeepSeek V4 Pro | 57.97 | #196 |
| 6 | DeepSeek V3.2 | 63.31 | #419 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=litmus-code-execution · How It Works · Data refreshed daily, snapshot 2026-11-21.