Long Video Understanding on LongVideoBench 473s
67.5AccuracyMolmo2
Evaluation Results
| Method | Links | |
|---|---|---|
| Molmo2Size=8B, Tokens per frame=832026.04 | 67.5 | |
| GPT-4oSize=-, Tokens per frame=-2026.04 | 66.7 | |
| Tempo*Size=6B, Tokens per frame=0.5–16, Visual Budget=8K, Actual tokens/frame=3.12026.04 | 65.1 | |
| VideoChat-FlashSize=7B, Tokens per frame=162026.04 | 64.7 | |
| Tempo*Size=6B, Tokens per frame=0.5–16, Visual Budget=4K, Actual tokens/frame=2.82026.04 | 64.5 | |
| Gemini 1.5 ProSize=-, Tokens per frame=-2026.04 | 64 | |
| InternVL3.5Size=8B, Tokens per frame=2562026.04 | 62.1 | |
| StormSize=7B, Tokens per frame=642026.04 | 60.5 | |
| VideoLLaMA3*Size=7B, Tokens per frame=≤ 912026.04 | 59.8 | |
| BIMBASize=7B, Tokens per frame=362026.04 | 59.5 | |
| LLaVA-VideoSize=7B, Tokens per frame=6762026.04 | 58.2 | |
| LongVILASize=7B, Tokens per frame=1962026.04 | 57.1 | |
| LLaVA-OneVisionSize=7B, Tokens per frame=1962026.04 | 56.4 | |
| Qwen2.5-VLSize=7B, Tokens per frame=19242026.04 | 56 | |
| KangarooSize=8B, Tokens per frame=2562026.04 | 54.8 | |
| LongLLaVASize=A13B, Tokens per frame=1442026.04 | 53.5 |