OccuBench (Mixed Faults): leaderboard
Metric: Completion rate (%) over all 382 tasks with about half explicit and half implicit injected faults (E3, two transient fault events of two calls each); 382 tool-use task instances from 100 professional scenarios, the environment simulated by Gemini-3-Flash-Preview and each trajectory passed or failed by a rubric verifier; thinking mode on, high reasoning effort where configurable; higher is better. Source: arxiv.org. Saturation forecast: Around January 2027. 9 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.2 (High) | 67 |
| 2 | Gemini 3.1 Pro (Preview) | 65.2 |
| 3 | Claude Opus 4.6 (High) | 63.9 |
| 4 | Qwen 3.5 Plus (Thinking) | 54.2 |
| 5 | Claude Sonnet 4.6 (Thinking) | 52.9 |
| 6 | DeepSeek V3.2 (Thinking) | 51.6 |
| 7 | GLM-5 (Thinking) | 47.4 |
| 8 | MiniMax-M2.7 | 46.9 |
| 9 | Kimi K2.5 (Thinking) | 40.1 |
Interactive version: theaggregate.ai/benchmark?slug=occubench-mixed-faults · How It Works · Data refreshed daily, snapshot 2026-10-07.