llava-v1.6-vicuna-7B-hf: benchmark results
Provider: LLaVA Team. Access: Open.
Aggregate Rating 1353 ± 47, rank #7597 of 10489 rated models, from 37 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| CoreCognition Image only Perspective | 49.14 | Accuracy (%) | 97.4 |
| UVLM Street-View Benchmark (Standard) | 83.1 | Mean proximity score (%) in standard mode (direct numeric an | 83.3 |
| CoreCognition Image only Intuitivephysics | 60.06 | Accuracy (%) | 78.3 |
| CoreCognition Image only Permanence | 62.5 | Accuracy (%) | 67.8 |
| UVLM Street-View Benchmark (Reasoning) | 77 | Mean proximity score (%) in advanced-reasoning mode (chain o | 64.3 |
| CoreCognition Image only Spatiality | 47.54 | Accuracy (%) | 31.6 |
| VRIQ (Abstract) - Sequence Completion | 23.5 | Accuracy (%) on the 200 abstract-domain VRIQ sequence comple | 29.2 |
| CoreCognition Image only Conservation | 36.51 | Accuracy (%) | 28.9 |
| CoreCognition Image only Boundary | 51.4 | Accuracy (%) | 18.4 |
| CoreCognition Image only Mean | 42.94 | Accuracy (%) | 18.4 |
| CoreCognition Image only Perceptualconstancy | 48.72 | Accuracy (%) | 18.4 |
| CoreCognition Image only Mechanicalreasoning | 41.76 | Accuracy (%) | 17.1 |
Interactive version: theaggregate.ai/model?slug=llava-v1-6-vicuna-7b-hf · How It Works · Data refreshed daily, snapshot 2026-11-20.