General Video Understanding on LongVideoBench
60.5AccuracyLVR-7B
Evaluation Results
| Method | Links | |
|---|---|---|
| LVR-7BFrames=32, Architecture Category=Latent MLLM2026.06 | 60.5 | |
| LVR-7BFrames=64, Architecture Category=Latent MLLM2026.06 | 60.1 | |
| VideoLatent-7B (Ours)Frames=64, Architecture Category=Latent MLLM2026.06 | 60.1 | |
| Qwen2.5-VL-7BFrames=64, Architecture Category=Standard MLLM2026.06 | 59.8 | |
| Open-o3-Video-7BFrames=64, Architecture Category=Standard MLLM2026.06 | 59.7 | |
| Qwen2.5-VL-7BFrames=32, Architecture Category=Standard MLLM2026.06 | 58.8 | |
| Open-o3-Video-7BFrames=32, Architecture Category=Standard MLLM2026.06 | 58.8 | |
| VideoLatent-7B (Ours)Frames=32, Architecture Category=Latent MLLM2026.06 | 58.3 | |
| Mull-Token-7BFrames=64, Architecture Category=Latent MLLM2026.06 | 58.3 | |
| Video-R1-7BFrames=64, Architecture Category=Standard MLLM2026.06 | 57.6 | |
| VideoLatent-7B (Ours)Frames=16, Architecture Category=Latent MLLM2026.06 | 57.3 | |
| Mull-Token-7BFrames=32, Architecture Category=Latent MLLM2026.06 | 57.2 | |
| Video-R1-7BFrames=32, Architecture Category=Standard MLLM2026.06 | 56.8 | |
| LVR-7BFrames=16, Architecture Category=Latent MLLM2026.06 | 56.4 | |
| Video-R1-7BFrames=16, Architecture Category=Standard MLLM2026.06 | 55.8 | |
| Mull-Token-7BFrames=16, Architecture Category=Latent MLLM2026.06 | 55.7 | |
| Qwen2.5-VL-7BFrames=16, Architecture Category=Standard MLLM2026.06 | 55.4 | |
| MECoT2025.12 | 52.13 | |
| LV-OV2025.12 | 51.91 | |
| DeepSportFrames=12.72025.11 | 45.9 | |
| Qwen2.5-VL-7B-InstructFrames=322025.11 | 43.2 | |
| Monet-7BFrames=64, Architecture Category=Latent MLLM2026.06 | 19.8 | |
| Monet-7BFrames=16, Architecture Category=Latent MLLM2026.06 | 19.7 | |
| Monet-7BFrames=32, Architecture Category=Latent MLLM2026.06 | 18.3 |