| Paper | Code | Accuracy | ModelName | ReleaseDate |
|---|---|---|---|---|
| V-JEPA 2: Self-Supervised Video Models Enable Understanding, Prediction and Planning | ✓ Link | 76.9 | V-JEPA 2 ViT-g (1B, 384px) + Llama 3.1 (8B) | 2025-06-11 |
| VideoChat3: Fully Open Video MLLM for Efficient and Generalist Video Understanding | ✓ Link | 75.6 | VideoChat3-4B | 2026-07-16 |
| LLaVA-OneVision-2: Towards Next-Generation Perceptual Intelligence | ✓ Link | 74.5 | LLaVA-OneVision-2-8B | 2026-05-25 |
| Before It Fades: Reinforcing Temporal Representations at Inference Time in VideoLLMs | ✓ Link | 74.0 | Qwen3-VL-8B + TAI | 2026-10-01 |
| Molmo2: Open Weights and Data for Vision-Language Models with Video Understanding and Grounding | ✓ Link | 73.4 | Molmo2-8B | 2026-01-15 |
| PerceptionLM: Open-Access Data and Models for Detailed Visual Understanding | ✓ Link | 72.7 | PLM-8B | 2025-04-17 |
| CoPE-VideoLM: Codec Primitives For Efficient Video Language Models | 68.9 | CoPE 7B | 2026-02-13 |