StreamPro-Bench - Perceptual Understanding: leaderboard
Metric: StreamPro-F1 (%; a response counts only when a Gemini 2.5 Pro judge credits its content and it arrives inside the task time window, and precision and recall over the response trajectory penalize extra and missed responses; videos streamed at 1 FPS; event understanding, object understanding and anomaly alert, weighted as in the paper). Source: arxiv.org. Saturation forecast: Around November 2027. 11 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | MiniCPM-o-4.5 | 13.5 |
| 2 | Qwen 3 VL 8B | 6.6 |
| 3 | Qwen 2.5 VL 7B | 3.4 |
Interactive version: theaggregate.ai/benchmark?slug=streampro-bench-perceptual-understanding · How It Works · Data refreshed daily, snapshot 2026-09-26.