Short Video Understanding on MVBench
76.4AccuracyInternVL-2.5
Evaluation Results
| Method | Links | |
|---|---|---|
| InternVL-2.5Evaluation frame rate=2 FPS2025.05 | 76.4 | |
| Seed1.5-VLThinking mode=true, Evaluation frame rate=2 FPS2025.05 | 74.4 | |
| Seed1.5-VLThinking mode=false, Evaluation frame rate=2 FPS2025.05 | 74.3 | |
| AdaSparkModel Size=Mid Size, Backbone=Qwen-2.5-VL-7B, SFT=true2026.04 | 70.3 | |
| VideoChat-Flash-2BModel Size=Small Size2026.04 | 70 | |
| Qwen-2.5-VL-7BModel Size=Mid Size, SFT=false2026.04 | 69.6 | |
| InternVL2.5-2BModel Size=Small Size2026.04 | 68.8 | |
| MoBAModel Size=Mid Size, Backbone=Qwen-2.5-VL-7B, SFT=true2026.04 | 67.5 | |
| ToMeModel Size=Mid Size, Backbone=Qwen-2.5-VL-7B, SFT=true2026.04 | 67.2 | |
| Qwen-2.5-VL-7B + SFTModel Size=Mid Size, SFT=true2026.04 | 67 | |
| FastVModel Size=Mid Size, Backbone=Qwen-2.5-VL-7B, SFT=true2026.04 | 67 | |
| FrameFusionModel Size=Mid Size, Backbone=Qwen-2.5-VL-7B, SFT=true2026.04 | 66.8 | |
| AdaSparkModel Size=Small Size, Backbone=Qwen-2.5-VL-3B, SFT=true2026.04 | 66.5 | |
| Qwen-2.5-VL-3BModel Size=Small Size, SFT=false2026.04 | 65.7 | |
| FastVModel Size=Small Size, Backbone=Qwen-2.5-VL-3B, SFT=true2026.04 | 65.6 | |
| FrameFusionModel Size=Small Size, Backbone=Qwen-2.5-VL-3B, SFT=true2026.04 | 65.4 | |
| Qwen-2.5-VL-3B + SFTModel Size=Small Size, SFT=true2026.04 | 64.9 | |
| ToMeModel Size=Small Size, Backbone=Qwen-2.5-VL-3B, SFT=true2026.04 | 64.5 | |
| MoBAModel Size=Small Size, Backbone=Qwen-2.5-VL-3B, SFT=true2026.04 | 64 | |
| FreshMembackbone=Qwen2VL-7B2026.02 | 63.8 | |
| VideoChat2-7BModel Size=Mid Size2026.04 | 62.3 | |
| LongVU-3BModel Size=Small Size2026.04 | 60.9 | |
| Qwen2VL-7B2026.02 | 59.1 | |
| LLaVA-Video-7B2026.02 | 58.6 | |
| LLaVA-OneVision-7B2026.02 | 56.7 | |
| Video-XL-7BModel Size=Mid Size2026.04 | 55.3 | |
| MovieChat-7B2026.02 | 55.1 | |
| LLaMA-VID-7BModel Size=Mid Size2026.04 | 41.4 |