llava-onevision-qwen2-7B-ov-hf: benchmark results
Provider: LLaVA Team. Access: Open.
Aggregate Rating 1503 ± 40, rank #2194 of 10870 rated models, from 23 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Modality Arbitration - MC2 - Aligned Image Accuracy | 95.85 | Score (%) | 100 |
| Modality Arbitration - MC2 - Visual-only Image Accuracy | 86.8 | Score (%) | 100 |
| Modality Arbitration - MMCD - Base Image-Target Accuracy | 97.64 | Score (%) | 100 |
| Modality Arbitration - MMCD - CoT Image-Target Accuracy | 97.3 | Score (%) | 100 |
| MUN-vis (5-Shot) | 47.4 | Win rate (%; share of MUN-vis items where a GPT-4o judge, Al | 83.3 |
| Modality Arbitration - MC2 - Base Image-Target Accuracy | 33.45 | Score (%) | 66.7 |
| Modality Arbitration - MC2 - Conflicting Image Accuracy | 33.45 | Score (%) | 66.7 |
| MUN-vis (Zero-Shot) | 30.1 | Win rate (%; share of MUN-vis items where a GPT-4o judge, Al | 50 |
| Modality Arbitration - CorrelationQA - Base Image-Target Accuracy | 17.82 | Score (%) | 33.3 |
| Modality Arbitration - CorrelationQA - CoT Image-Target Accuracy | 19.12 | Score (%) | 33.3 |
| Modality Arbitration - MC2 - Conflicting Text Accuracy | 54.55 | Score (%) | 33.3 |
| Modality Arbitration - MC2 - Text-only Text Accuracy | 87.05 | Score (%) | 33.3 |
Interactive version: theaggregate.ai/model?slug=llava-onevision-qwen2-7b-ov-hf · How It Works · Data refreshed daily, snapshot 2026-11-12.