Long Video Understanding on Video-MME LVBench HippoVlog long
70.1Average AccuracyGPT-5
Evaluation results
| Method | Links | |
|---|---|---|
| GPT-52026.07 | 70.1 | |
| Light-OmniSize=7B2026.07 | 64.8 | |
| WorldMM-8BSize=8B2026.07 | 63.7 | |
| Gemini 2.5 Pro2026.07 | 61.6 | |
| Qwen3-VL-8BSize=8B2026.07 | 61.2 | |
| M3-AgentSize=32B2026.07 | 58.9 | |
| Qwen2.5-Omni-7BSize=7B2026.07 | 53 | |
| Video-RAG2026.07 | 51.2 | |
| HippoMM2026.07 | 50.6 | |
| Ego-R1Size=3B2026.07 | 45.2 | |
| LightRAG2026.07 | 41.5 |