MRR-Benchmark: leaderboard
The Multi-modal Robustness (MMR) benchmark, evaluating whether vision-language models correctly understand image content and resist misleading or leading questions designed to induce incorrect answers.
Metric: Total Column Score. Source: huggingface.co. Saturation forecast: Estimated already saturated. 14 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude 3.5 Sonnet | 463 | #350 |
| 2 | GPT-4o | 457 | #343 |
| 3 | InternVL2-8B | 368 | #817 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=mrr-benchmark · How It Works · Data refreshed daily, snapshot 2026-10-20.