Video Question Answering on Video-MME long overall durations
66.46Acc (Long, -subs)Video-QTR
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| Video-QTR2025.12 | 66.46 | 75.19 | 72.59 | 80.93 | — | |
| gemini2.5-pro2025.12 | 61.49 | 65 | 70.53 | 76.81 | — | |
| qwen2.5-vl-max2025.12 | 51.47 | 60.37 | 65.96 | 72.99 | — | |
| VILA-1.52025.12 | 50.8 | 52 | 59 | 59.4 | — | |
| gpt-4.12025.12 | 50.51 | 61.04 | 67.23 | 73.9 | — | |
| Flash-Vstream2025.12 | 50.3 | 61.4 | 61.2 | 67 | — | |
| VITA-1.02025.12 | 48.6 | 50.9 | 55.8 | 59.2 | — | |
| VideoLLaMA2025.12 | 47.6 | 49 | 52.4 | 54.71 | — | |
| VideoChat2025.12 | 45.6 | 53.3 | 55.3 | 59.7 | — | |
| Video-LLaVA2025.12 | 36.2 | 38.1 | 39.9 | 41.6 | — | |
| MovieChat2025.12 | 33.4 | 36.13 | 33.4 | 37.11 | — | |
| VideoChat2-Mistral2025.12 | 33.2 | 39.2 | 39.5 | 43.8 | — | |
| ST-LLM2025.12 | 31.3 | 36.9 | 37.9 | 42.3 | — | |
| Dispider-7B# Frames=1 fps2026.03 | — | — | — | — | 57.2 | |
| Kangaroo-7B# Frames=642026.03 | — | — | — | — | 56 | |
| LLaVA-Next-Video-7B# Frames=322026.03 | — | — | — | — | 46.6 | |
| LongVA-7B# Frames=1282026.03 | — | — | — | — | 52.6 | |
| MovieChat-7B# Frames=20482026.03 | — | — | — | — | 38.2 | |
| Qwen2.5-VL-7B# Frames=1 fps2026.03 | — | — | — | — | 63.2 | |
| TimeChat-Online-7B# Frames=1 fps2026.03 | — | — | — | — | 62.5 | |
| Video-CCAM-14B# Frames=962026.03 | — | — | — | — | 53.2 | |
| VideoChat-Online-8B# Frames=2 fps2026.03 | — | — | — | — | 52.8 | |
| VideoChat2-7B# Frames=162026.03 | — | — | — | — | 39.5 | |
| VideoXL-7B# Frames=1282026.03 | — | — | — | — | 55.5 | |
| WAT# Frames=1 fps2026.03 | — | — | — | — | 62.4 |