Video Multi-modal Evaluation on VideoMME
66.9Overall ScoreQwen3-VL-8B-Instruct
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| Qwen3-VL-8B-InstructMax Input Frames=642026.03 | 66.9 | 78.8 | 66.2 | 55.8 | |
| VidCom2Max Input Frames=64, Retention Ratio=35%2026.03 | 66 | 77.3 | 64.9 | 55.8 | |
| V-CASTMax Input Frames=64, Retention Ratio=35%2026.03 | 66 | 77.6 | 65.6 | 55 | |
| Qwen3-VL-8B-InstructMax Input Frames=322026.03 | 64.5 | 76 | 60.4 | 57 | |
| VisionZipMax Input Frames=64, Retention Ratio=35%2026.03 | 64.3 | 74.9 | 62.2 | 55.9 | |
| HoliTomMax Input Frames=64, Retention Ratio=35%2026.03 | 64 | 75.8 | 61.8 | 54.3 | |
| V-CASTMax Input Frames=32, Retention Ratio=35%2026.03 | 63.9 | 74.7 | 60.8 | 56.3 | |
| FastVIDMax Input Frames=64, Retention Ratio=35%2026.03 | 63 | 74.2 | 60.3 | 54.4 | |
| VidCom2Max Input Frames=32, Retention Ratio=35%2026.03 | 62.9 | 73.6 | 59.9 | 55.1 | |
| FlashVIDMax Input Frames=32, Retention Ratio=35%2026.03 | 62.7 | 75.2 | 59.6 | 53.3 | |
| FastVIDMax Input Frames=32, Retention Ratio=35%2026.03 | 61.9 | 73.1 | 59.1 | 53.6 | |
| HoliTomMax Input Frames=32, Retention Ratio=35%2026.03 | 61.3 | 71.9 | 57.9 | 54 | |
| VisionZipMax Input Frames=32, Retention Ratio=35%2026.03 | 60.8 | 70.9 | 57 | 54.6 |