| video-question-answering-on-ovo-bench | Qwen2.5-VL-7B | – | Real-Time Visual Perception: 59.9Backward Tracing: 44.7 |
| video-question-answering-on-ovo-bench | SimpleStream (Qwen2.5-VL-7B, 2 frames) | – | Real-Time Visual Perception: 75.0Backward Tracing: 49.5 |
| video-question-answering-on-ovo-bench | SimpleStream (Qwen2.5-VL-7B, 4 frames) | – | Real-Time Visual Perception: 78.4Backward Tracing: 51.9 |
| video-question-answering-on-ovo-bench | SimpleStream (Qwen2.5-VL-7B, 8 frames) | – | Real-Time Visual Perception: 76.6Backward Tracing: 50.8 |
| video-question-answering-on-ovo-bench | SimpleStream (Qwen3-VL-8B, 2 frames) | – | Real-Time Visual Perception: 79.3Backward Tracing: 53.5 |
| video-question-answering-on-ovo-bench | SimpleStream (Qwen3-VL-8B, 4 frames) | – | Real-Time Visual Perception: 81.4Backward Tracing: 54.0 |
| video-question-answering-on-ovo-bench | SimpleStream (Qwen3-VL-8B, 8 frames) | – | Real-Time Visual Perception: 79.9Backward Tracing: 54.9 |