🎥 LVU VLM Leaderboard
Benchmarks: Video-MME, MLVU, LVBench, LongVideoBench
Sort by
| Rank | Model | Video-MME | MLVU | LongVideoBench | LVBench | Size | Category |
|---|---|---|---|---|---|---|---|
| 1 | Gemini-1.5-Pro | 75.0 | 62.9 | 64.0 | 33.1 | - | Proprietary Models |
| 2 | GPT-4o | 71.9 | 64.6 | 66.7 | 34.7 | - | Proprietary Models |
| 3 | Qwen2-VL | 63.3 | 64.2 | 52.4 | 42.0 | 7B | Open-source Transformer-based LMMs |
| 4 | GPT-4V | 59.9 | 49.2 | 59.1 | - | - | Proprietary Models |
| 5 | LLaVA-OneVision | 58.2 | 62.6 | 56.4 | - | 7B | Open-source Transformer-based LMMs |
| 6 | InternVL2 | 58.2 | 48.1 | 51.8 | - | 8B | Open-source Transformer-based LMMs |
| 7 | VAMBA | 57.8 | 65.9 | 55.9 | 42.1 | 10B | Open-source Efficient LMMs |
| 8 | Kangaroo | 56.0 | 61.0 | 54.8 | - | 8B | Open-source Transformer-based LMMs |
| 9 | Video-XL | 55.5 | 64.9 | 50.7 | 36.8 | 7B | Open-source Efficient LMMs |
| 10 | LongVU | 55.3 | 65.4 | 53.5 | 37.8 | 7B | Open-source Efficient LMMs |
| 11 | Phi-4-Mini | 55.0 | 60.1 | 46.7 | - | 5.6B | Open-source Transformer-based LMMs |
| 12 | LongVA | 52.4 | 56.3 | 51.8 | - | 7B | Open-source Transformer-based LMMs |
| 13 | LongLLaVA | 51.6 | 53.3 | 42.1 | 31.2 | 9B | Open-source Efficient LMMs |
| 14 | Video-CCAM | 50.3 | 58.5 | - | - | 9B | Open-source Transformer-based LMMs |
| 15 | LLaVA-Mini | 40.3 | 44.3 | 19.3 | 17.6 | 7B | Open-source Efficient LMMs |
| 16 | ShareGPT4Video | 39.9 | 46.4 | 39.7 | - | 7B | Open-source Transformer-based LMMs |
| 17 | VideoChat2 | 39.5 | 47.9 | 39.3 | - | 7B | Open-source Transformer-based LMMs |