LLaVA-v1.6-Vicuna-13B: benchmark results
Provider: LLaVA Team. Access: Open.
Aggregate Rating 1416 ± 33, rank #5336 of 10916 rated models, from 55 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| GeoHeight-Bench - Elevation Retrieval | 55.64 | Text accuracy (%) on the Elevation Retrieval task (the model | 100 |
| GeoHeight-Bench | 22.5 | Text accuracy (%), the mean of the 7 task accuracies, on Geo | 91.7 |
| GeoHeight-Bench+ - Pixel Information | 12.62 | Text accuracy (%) on the Pixel Information task (the model g | 91.7 |
| GeoHeight-Bench - Pixel Information | 9.48 | Text accuracy (%) on the Pixel Information task (the model g | 79.2 |
| GeoHeight-Bench - Height Ranking | 7.33 | Text accuracy (%) on the Height Ranking task (the model give | 75 |
| GeoHeight-Bench+ - Slope Information | 15.54 | Text accuracy (%) on the Slope Information task (the model g | 75 |
| GeoHeight-Bench+ - Threshold Segmentation | 38.07 | Text accuracy (%) on the Threshold Segmentation task (the mo | 75 |
| GeoHeight-Bench - Threshold Segmentation | 44.98 | Text accuracy (%) on the Threshold Segmentation task (the mo | 66.7 |
| Epoch AI - Science Qa | 73.6 | Score | 65.2 |
| GeoHeight-Bench+ - Instance Extraction | 6.71 | Text accuracy (%) on the Instance Extraction task (the model | 62.5 |
| AgroTools - Summary Accuracy | 21.83 | Summary accuracy (0-100): normalized similarity of the final | 58.3 |
| GeoHeight-Bench - Pattern Description | 20.4 | Text accuracy (%) on the Pattern Description task (the model | 58.3 |
Interactive version: theaggregate.ai/model?slug=llava-v1-6-vicuna-13b · How It Works · Data refreshed daily, snapshot 2026-11-08.