VideoChat-R1-7B: benchmark results
Provider: Other.
Aggregate Rating 1600 ± 43, rank #1017 of 10916 rated models, from 16 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| OVIBench (Multiple Choice) - Cancellation | 52.87 | Answer-selection accuracy (%; choosing, among four candidate | 80 |
| VERHallu - Causal QA | 40.8 | Accuracy (%; 497 seven-option questions on why an event happ | 79.4 |
| VERHallu - Subevent QA | 38.4 | Accuracy (%; 258 seven-option questions on which event occur | 76.5 |
| VERHallu - Temporal QA | 38.7 | Accuracy (%; 212 seven-option questions on what happens afte | 76.5 |
| MMVIAD | 43.3 | Mean of three accuracies and visible-time mIoU (%; MMVIAD-St | 70 |
| MMVIAD - Defect Classification | 32.1 | Accuracy (%; MMVIAD-Standard test split, 1,101 rendered 2-se | 70 |
| MMVIAD - Object Classification | 68.8 | Accuracy (%; MMVIAD-Standard test split, 1,101 rendered 2-se | 70 |
| OVIBench (Multiple Choice) - Correction | 89.58 | Answer-selection accuracy (%; choosing, among four candidate | 60 |
| OVIBench (Multiple Choice) - False Trigger | 56.54 | Answer-selection accuracy (%; choosing, among four candidate | 60 |
| MMVIAD - Anomaly Detection | 41.1 | Accuracy (%; MMVIAD-Standard test split, 1,101 rendered 2-se | 50 |
| MMVIAD - Anomaly Visible-Time Localization | 31.1 | Mean temporal IoU (%; MMVIAD-Standard test split, predicted | 45 |
| MMVIAD - Unseen Categories | 45.4 | Mean of three accuracies and visible-time mIoU (%; MMVIAD-Un | 40 |
Interactive version: theaggregate.ai/model?slug=videochat-r1-7b · How It Works · Data refreshed daily, snapshot 2026-11-08.