H2OVL-Mississippi-2B: benchmark results
Provider: H2O.ai.
Aggregate Rating 1427 ± 47, rank #4839 of 10916 rated models, from 6 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| MA-Bench - Coarse-Grained Recognition | 17.7 | Accuracy (%) on the coarse-grained (body-level) micro-action | 38.1 |
| MA-Bench (Open-Ended) | 0.11 | GPT-4o judge score (0 to 5) averaged over the six open-ended | 28.6 |
| VRBench | 47.15 | Overall (imported) | 7.1 |
| MA-Bench (Closed-Ended) | 39.42 | Accuracy (%) averaged over the six closed-ended tasks (coars | 4.8 |
| MA-Bench - Fine-Grained Recognition | 19.2 | Accuracy (%) on the fine-grained (action-level) micro-action | 4.8 |
| MMVU | 28.8 | Hidden-test overall accuracy, chain-of-thought (%) | 1.4 |
Interactive version: theaggregate.ai/model?slug=h2ovl-mississippi-2b · How It Works · Data refreshed daily, snapshot 2026-11-08.