LLaVA-Med-7B: benchmark results
Provider: LLaVA Team.
Aggregate Rating 1407 ± 42, rank #5729 of 10916 rated models, from 32 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| SurgCoT - Causal Action Ordering | 89.95 | Causal action ordering main-question accuracy (%) on SurgCoT | 77.8 |
| SurgCoT - Affordance Mapping | 62.29 | Affordance mapping main-question accuracy (%) on SurgCoT sur | 66.7 |
| SurgCoT - Micro-Transition Localization | 71.43 | Micro-transition localization main-question accuracy (%) on | 44.4 |
| Medical MM Leaderboard - PathVQA | 38.8 | Accuracy (%) | 23.8 |
| SurgCoT | 68.15 | Average over the five reasoning tasks of main-question accur | 22.2 |
| SurgCoT - Cue-Action Alignment | 66.1 | Cue-action alignment main-question accuracy (%) on SurgCoT s | 22.2 |
| X-PCR - Lesion Characterization | 55.57 | Accuracy (%) at stage 3, characterizing lesion morphology an | 20 |
| X-PCR - Disease Diagnosis | 46.18 | Accuracy (%) at stage 4, diagnosing the disease from the fin | 15 |
| Scale-VQA - Confirmation | 16.92 | Accuracy (%; multiple-choice questions on paired 10x/40x/200 | 11.1 |
| SurgCoT - Anomaly Onset Tracking | 50.97 | Anomaly onset tracking main-question accuracy (%) on SurgCoT | 11.1 |
| Medical MM Leaderboard - OMVQA | 44.3 | Accuracy (%) | 10.5 |
| Medical MM Leaderboard - MMMU-Med | 29.3 | Accuracy (%) | 9.5 |
Interactive version: theaggregate.ai/model?slug=llava-med-7b · How It Works · Data refreshed daily, snapshot 2026-11-08.