LLaVA-OneVision-Qwen2-7B: benchmark results
Provider: LLaVA Team.
Aggregate Rating 1489 ± 25, rank #2527 of 10916 rated models, from 59 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| MVPBench - Multi-View Perception Pairing | 72.04 | MVPBench normalized proficiency score (accuracy minus chance | 96.4 |
| Vinoground (Text Score) | 41.6 | Text Score | 82.6 |
| Vinoground (Group Score) | 14.6 | Group Score | 78.3 |
| Vinoground (Video Score) | 29.4 | Video Score | 78.3 |
| MVPBench - Caption-Free Temporal Splicing | -0.73 | MVPBench normalized proficiency score (accuracy minus chance | 71.4 |
| MVPBench - Caption-Guided Temporal Splicing | 13.92 | MVPBench normalized proficiency score (accuracy minus chance | 71.4 |
| MVPBench - Video Clarity Ranking | 31.43 | MVPBench normalized proficiency score (accuracy minus chance | 71.4 |
| MVPBench - Professional Action Evaluation | 8.58 | MVPBench normalized proficiency score (accuracy minus chance | 64.3 |
| TempCompass | 64.22 | Avg All (%) | 64.3 |
| MVPBench - Daily Action Evaluation | 33.34 | MVPBench normalized proficiency score (accuracy minus chance | 60.7 |
| MVPBench - Gait Recognition Matching | 17.5 | MVPBench normalized proficiency score (accuracy minus chance | 60.7 |
| MVPBench - Olympic Sports Matching | 4 | MVPBench normalized proficiency score (accuracy minus chance | 60.7 |
Interactive version: theaggregate.ai/model?slug=llava-onevision-qwen2-7b · How It Works · Data refreshed daily, snapshot 2026-11-08.