FailBench - ur5fail: leaderboard
FailBench evaluates video-language models and specialist detectors judging robot task success across real and simulated robot videos under a shared prompt, input and decoding recipe. Macro results average two-class datasets; the failure-only reflect subset uses failure recall. This series reports FailBench - ur5fail from 2609.03611v1, S4.T2: Balanced Accuracy (%). Higher is better.
Metric: Balanced Accuracy (%). Source: arxiv.org. Saturation forecast: Around February 2027. 13 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3 Flash | 84 | #159 |
| 2 | Gemma 4 31B (IT) | 79 | #343 |
| 3 | GPT-4o | 76 | #741 |
| 4 | Qwen 3 VL 4B (Thinking) | 66 | #1303 |
| 5 | Qwen 3 VL 8B (Thinking) | 62 | #887 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=failbench-ur5fail · How It Works · Data refreshed daily, snapshot 2026-11-12.