Multi-choice Video Question Answering on NExT-QA
83.2Overall AccuracyLLaVA-Video-7B
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| LLaVA-Video-7B2025.07 | 83.2 | — | — | — | |
| VideoLLaMA3-7BFrames=162025.10 | 82.3 | — | — | — | |
| FrameOracleBackbone=VideoLLaMA3-7B, Frames=64→13.92025.10 | 82 | — | — | — | |
| FrameOracleBackbone=LLaVA-Video-7B, Frames=64→13.92025.10 | 81.6 | — | — | — | |
| FrameOracleBackbone=VideoLLaMA3-7B, Frames=16→10.42025.10 | 81.2 | — | — | — | |
| LLaVA-Video-7BFrames=162025.10 | 81 | — | — | — | |
| FrameOracleBackbone=LLaVA-Video-7B, Frames=16→10.42025.10 | 80.4 | — | — | — | |
| InternVL2.5-8B-ITG-32Frames Sampled=322025.07 | 79.5 | — | — | — | |
| LLaVA-OV-7B2025.07 | 79.4 | — | — | — | |
| mPLUG-Owl3-8B2025.07 | 78.6 | — | — | — | |
| FrameOracleBackbone=LLaVA-OneVision-7B, Frames=64→13.92025.10 | 78.5 | — | — | — | |
| LLaVA-OneVision-7BFrames=162025.10 | 78.2 | — | — | — | |
| FrameOracleBackbone=LLaVA-OneVision-7B, Frames=16→10.42025.10 | 77.6 | — | — | — | |
| FrameOracleBackbone=Qwen3-VL-8B, Frames=128→27.82025.10 | 77.3 | — | — | — | |
| Qwen3-VL-8BFrames=322025.10 | 76.6 | — | — | — | |
| FrameOracleBackbone=Qwen2.5-VL-3B, Frames=128→27.82025.10 | 76.1 | — | — | — | |
| FrameOracleBackbone=Qwen3-VL-8B, Frames=32→20.92025.10 | 76.1 | — | — | — | |
| Qwen2.5-VL-3BFrames=322025.10 | 75.4 | — | — | — | |
| FrameOracleBackbone=Qwen2.5-VL-3B, Frames=32→20.92025.10 | 74.8 | — | — | — | |
| Qwen2.5-VL-7B2025.07 | 70.2 | — | — | — | |
| LongVA-7B2025.07 | 68.3 | — | — | — | |
| VILA-40B2025.07 | 67.9 | — | — | — | |
| MAMA-VGTBackbone=VGT2026.01 | 66.3 | 72.7 | 70.6 | 62.2 | |
| MAMA-VIOLETBackbone=VIOLET2026.01 | 65.8 | 71.2 | 68.6 | 61.1 | |
| LaViLa-VGTBackbone=VGT2026.01 | 64.8 | 70.9 | 69.2 | 59.3 | |
| Vid2Seq-VGTBackbone=VGT2026.01 | 64.2 | 70.3 | 67 | 58.5 | |
| VGT2026.01 | 61.8 | 69.6 | 65.4 | 56.2 | |
| OmniVL2026.01 | 59.9 | 67.1 | 63.8 | 55.9 | |
| VindLU2026.01 | 59.8 | 68.4 | 59.7 | 55.7 | |
| MERLOT2026.01 | 59.4 | 66.6 | 62.5 | 58.2 | |
| VIOLET2026.01 | 58.5 | 67.7 | 58 | 50.7 | |
| All-in-one2026.01 | 57.9 | 64.8 | 63.9 | 50.9 | |
| LAVENDER2026.01 | 56.7 | 63.4 | 56.5 | 54.6 | |
| Singularity2026.01 | 54.8 | 60 | 61.1 | 50.9 | |
| VideoLLaMA2-7BFrames=162025.10 | 45.4 | — | — | — |