Long-Video Understanding on VideoMME (1-60m, w/o subtitles)
75ScoreGemini-1.5-Pro
Evaluation results
| Method | Links | |
|---|---|---|
| Gemini-1.5-ProModel Type=Transformer, Size=–2026.02 | 75 | |
| Qwen2.5+FlexSelectModel Type=Transformer, Size=72B2026.02 | 74.4 | |
| GPT-4oModel Type=Transformer, Size=–2026.02 | 71.9 | |
| Nemotron-Nano-V2 VL + All layer reductionModel Type=Hybrid, Size=12B2026.02 | 69.7 | |
| LLaVA-Video+FlexSelectModel Type=Transformer, Size=7B2026.02 | 68.9 | |
| Qwen3 VL + All layer reductionModel Type=Transformer, Size=8B2026.02 | 68.52 | |
| Qwen2.5 VL+FlexSelectModel Type=Transformer, Size=7B2026.02 | 68.2 | |
| ViLAMPModel Type=Transformer, Size=7B2026.02 | 67.5 | |
| Oryx-1.5Model Type=Transformer, Size=34B2026.02 | 67.3 | |
| Qwen2.5 VL+SparseVILAModel Type=Transformer, Size=7B2026.02 | 66.3 | |
| VideoLLaMA3Model Type=Transformer, Size=7B2026.02 | 66.2 | |
| NVILAModel Type=Transformer, Size=8B2026.02 | 64.2 | |
| SF-LLaVA-1.5Model Type=Transformer, Size=7B2026.02 | 63.9 | |
| Qwen2-VLModel Type=Transformer, Size=7B2026.02 | 63.3 | |
| ApolloModel Type=Transformer, Size=7B2026.02 | 61.3 | |
| LongVUModel Type=Transformer, Size=7B2026.02 | 60.6 | |
| MLLM-FSModel Type=Transformer, Size=7B2026.02 | 58.7 | |
| LLaVA-OnevisionModel Type=Transformer, Size=7B2026.02 | 58.2 |