StreamingVLM: benchmark results
Provider: Other.
Aggregate Rating 1593 ± 48, rank #1085 of 10916 rated models, from 17 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| SVCBench - O1-Snap | 36.7 | Gaussian Precision Accuracy (0-100): per query point exp(-(p | 84.6 |
| EgoMemReason - Temporal Counting | 29 | Accuracy (%; 100 multiple-choice questions) | 45.8 |
| EgoMemReason - Activity Pattern Inference | 32 | Accuracy (%; 50 multiple-choice questions) | 41.7 |
| SVCBench - O2-Unique | 26 | Gaussian Precision Accuracy (0-100): per query point exp(-(p | 38.5 |
| SVCBench - E1-Transit | 13.8 | Gaussian Precision Accuracy (0-100): per query point exp(-(p | 30.8 |
| SVCBench - E2-Episode | 20.7 | Gaussian Precision Accuracy (0-100): per query point exp(-(p | 30.8 |
| EgoMemReason - Cumulative State Tracking | 25 | Accuracy (%; 100 multiple-choice questions) | 25 |
| SVCBench - E1-Action | 15.8 | Gaussian Precision Accuracy (0-100): per query point exp(-(p | 23.1 |
| SVCBench | 19.1 | GPA (Global Performance Average, %) | 19.2 |
| EgoMemReason - Event Linking | 20 | Accuracy (%; 100 multiple-choice questions) | 16.7 |
| EgoMemReason - Event Ordering | 21 | Accuracy (%; 100 multiple-choice questions) | 16.7 |
| SVCBench - O2-Gain | 14.1 | Gaussian Precision Accuracy (0-100): per query point exp(-(p | 15.4 |
Interactive version: theaggregate.ai/model?slug=streamingvlm · How It Works · Data refreshed daily, snapshot 2026-11-08.