LLaVA-v1.6-Vicuna-7B: benchmark results
Provider: LLaVA Team. Access: Open.
Aggregate Rating 1429 ± 29, rank #4687 of 10916 rated models, from 54 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| GeoHeight-Bench+ - Instance Extraction | 7.77 | Text accuracy (%) on the Instance Extraction task (the model | 100 |
| GeoHeight-Bench+ - Slope Information | 16.39 | Text accuracy (%) on the Slope Information task (the model g | 100 |
| SaLAD - Safe Queries | 99.79 | Accuracy rate (%; GPT-4o judge; 943 benign queries answered | 85.3 |
| GeoHeight-Bench - Elevation Retrieval | 34.07 | Text accuracy (%) on the Elevation Retrieval task (the model | 66.7 |
| GeoHeight-Bench - Pixel Information | 9.16 | Text accuracy (%) on the Pixel Information task (the model g | 66.7 |
| GeoHeight-Bench+ - Elevation Retrieval | 7.43 | Text accuracy (%) on the Elevation Retrieval task (the model | 66.7 |
| GeoHeight-Bench+ - Pixel Information | 10.59 | Text accuracy (%) on the Pixel Information task (the model g | 66.7 |
| GeoHeight-Bench - Height Ranking | 5.83 | Text accuracy (%) on the Height Ranking task (the model give | 58.3 |
| GeoHeight-Bench - Threshold Segmentation | 44.54 | Text accuracy (%) on the Threshold Segmentation task (the mo | 50 |
| GeoHeight-Bench+ | 14.38 | Text accuracy (%), the mean of the 8 task accuracies, on Geo | 50 |
| GeoHeight-Bench+ - Pattern Description | 18.08 | Text accuracy (%) on the Pattern Description task (the model | 50 |
| SenseBench - Single-Image Perception | 45.83 | Exact-Match Accuracy (%; mean of the whether, what and how q | 50 |
Interactive version: theaggregate.ai/model?slug=llava-v1-6-vicuna-7b · How It Works · Data refreshed daily, snapshot 2026-11-08.