LLaVA-Next-Vicuna-7B: benchmark results
Provider: LLaVA Team. Released 2024-01-30. Access: Open.
Aggregate Rating 1318 ± 29, rank #8369 of 10916 rated models, from 460 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| OpenVLM MMT-Bench - CIM | 80 | Score (%) | 98.8 |
| OpenVLM MMT-Bench - Single Object Tracking | 80 | Score (%) | 98.8 |
| OpenVLM MMT-Bench - Deepfake Detection | 70 | Score (%) | 98.3 |
| OpenVLM MMT-Bench - 3D Indoor Recognition | 50 | Score (%) | 96.4 |
| OpenVLM MMT-Bench - Geometrical Relativity | 75 | Score (%) | 96.1 |
| OpenVLM MMT-Bench - Point Tracking | 85 | Score (%) | 93.9 |
| OpenVLM MMT-Bench - MIA | 72.5 | Score (%) | 91.3 |
| ProactiveBench - ChangeIt | 36.6 | Trajectory accuracy (%) on the ambiguous ChangeIt video fram | 90.5 |
| OpenVLM MMT-Bench - Film and Television Recognition | 95 | Score (%) | 90 |
| OpenVLM A-Bench - Nouns as Adjectives | 83.7 | Accuracy (%) | 89.6 |
| OpenVLM MMT-Bench - Plant Recognition | 90 | Score (%) | 88.8 |
| OpenVLM MMT-Bench - Spot The Diff | 80 | Score (%) | 88.6 |
Interactive version: theaggregate.ai/model?slug=llava-next-vicuna-7b · How It Works · Data refreshed daily, snapshot 2026-11-08.