VideoLLaMA2-7B: benchmark results
Provider: Other. Access: Open.
Aggregate Rating 1514 ± 20, rank #2136 of 10916 rated models, from 87 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AVI-Bench - Audio Multi-Instance Classification | 36.46 | Task score (%) on audio multi-instance classification (518 c | 96 |
| AVI-Bench - Audio Sensation QA | 24.3 | Task score (%) on primitive audio sensation (502 multiple-ch | 88 |
| AVI-Bench - Audio-Visual Sensation QA | 21.65 | Task score (%) on primitive audio-visual sensation (388 mult | 86 |
| VidPair-Halluc - Multi-Choice F1 | 62.91 | Multiple-choice F1 (%) for assessing and ranking clip-level | 85.7 |
| AVI-Bench - Visual-Referenced Audio Hallucination | 70 | Task score (%) on visual-referenced audio hallucination (250 | 76 |
| SceneBench - Scene QA | 28.7 | Accuracy (%) on SceneBench's 3,311 scene qa items (questions | 72.2 |
| AVI-Bench - Perception | 32.45 | Stage score (%) for the perception stage: mean of the audio | 68 |
| AVI-Bench - Primitive Sensation | 20.34 | Stage score (%) for the primitive sensation extension (AVI-B | 68 |
| SceneBench - Scene QA (Audio) | 32 | Accuracy (%) on SceneBench's 4,097 scene qa (audio) items (s | 66.7 |
| MLVU-Dev - Generation | 3.99 | Generation score, GPT-4 judged (0-10) | 65.5 |
| VideoOdyssey-AV - Captioning | 46.8 | Multiple-choice accuracy (%) on the 47 captioning questions | 63.6 |
| VideoOdyssey-AV - Temporal Ordering | 26 | Multiple-choice accuracy (%) on the 50 temporal ordering que | 63.6 |
Interactive version: theaggregate.ai/model?slug=videollama2-7b · How It Works · Data refreshed daily, snapshot 2026-11-08.