LongVA: benchmark results
Provider: Other.
Aggregate Rating 1487 ± 36, rank #2550 of 10916 rated models, from 19 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| CG-Bench - Clue Grounding mIoU | 2.91 | Temporal Grounding mIoU (%) | 61.9 |
| SceneBench - Scene QA | 27.7 | Accuracy (%) on SceneBench's 3,311 scene qa items (questions | 58.3 |
| SceneBench - Clip QA | 54.3 | Accuracy (%) on SceneBench's 368 clip qa items (questions ab | 55.6 |
| SceneBench | 53.3 | Average accuracy (%) over SceneBench's six tasks (title and | 50 |
| SceneBench - Inverse VQA | 40.1 | Accuracy (%) on SceneBench's 485 inverse vqa items (choosing | 50 |
| CG-Bench - Accuracy@IoU | 1.32 | Accuracy@IoU (%) | 47.6 |
| SceneBench - Scene QA (Audio) | 29.5 | Accuracy (%) on SceneBench's 4,097 scene qa (audio) items (s | 44.4 |
| CG-Bench | 28.7 | Long-Video MCQ Accuracy (%) | 42.9 |
| CG-Bench - Clue MCQ | 42.6 | Clue-Interval MCQ Accuracy (%) | 42.9 |
| CG-Bench - Open-Ended | 24.2 | Open-Ended Accuracy (%) | 42.9 |
| SceneBench - Title Prediction | 96.8 | Accuracy (%) on SceneBench's 345 title prediction items (cho | 33.3 |
| SceneBench - Comment Prediction | 71.1 | Accuracy (%) on SceneBench's 297 comment prediction items (c | 27.8 |
Interactive version: theaggregate.ai/model?slug=longva · How It Works · Data refreshed daily, snapshot 2026-11-08.