VLM-RobustBench (MMMU-Pro) - Severe-Failure Rate: leaderboard
Metric: Severe-failure rate (%): share of the 133 corrupted settings: 42 severity-based corruptions in nine categories (noise, blur, weather, digital, geometric and others) at low, mid and high severity, plus 7 binary transforms such as vertical flip and colour inversion; direct-answer prompting in which accuracy drops by more than 10 percent of the model's clean accuracy, on a stratified 20 percent sample (532 questions) of the 10-option MMMU-Pro standard set; lower is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 9 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3 VL 4B Instruct | 3.8 | #506 |
| 2 | Molmo2-8B | 5.3 | #522 |
| 3 | Qwen 3 VL 8B Instruct | 7.5 | #401 |
| 4 | Qwen 3 VL 30B A3B Instruct | 12.8 | #365 |
| 5 | Gemma 3 12B (IT) | 27.1 | #655 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #8 Claude Fable 5.1.
Interactive version: theaggregate.ai/benchmark?slug=vlm-robustbench-mmmu-pro-severe-failure-rate · How It Works · Data refreshed daily, snapshot 2026-10-11.