| Qwen3.8-Max: A New Bar for Coding and Cowork | ✓ Link | 90.4 | Qwen3.8-Max | 2026-08-03 |
| Kimi K3: Open Frontier Intelligence | ✓ Link | 90.0 | Kimi K3 Max | 2026-07-27 |
| Qwen3.6 | ✓ Link | 87.7 | Qwen3.6-27B | 2026-04-21 |
| Qwen3.5: Towards Native Multimodal Agents | ✓ Link | 87.5 | Qwen3.5-397B-A17B (w/ subtitles) | 2026-02-16 |
| Kimi K2.5: Visual Agentic Intelligence | ✓ Link | 87.4 | Kimi K2.5 | 2026-02-02 |
| Qwen3.5: Towards Native Multimodal Agents | ✓ Link | 87.3 | Qwen3.5-122B-A10B (w/ subtitles) | 2026-02-16 |
| Qwen3.5: Towards Native Multimodal Agents | ✓ Link | 87.0 | Qwen3.5-27B (w/ subtitles) | 2026-02-16 |
| Qwen3.5: Towards Native Multimodal Agents | ✓ Link | 86.6 | Qwen3.5-35B-A3B (w/ subtitles) | 2026-02-16 |
| Qwen3.6 | ✓ Link | 86.6 | Qwen3.6-35B-A3B | 2026-04-21 |
| MiniMax M3: Frontier Coding, 1M Context, Native Multimodality — All in One Model | ✓ Link | 85.4 | MiniMax M3 (w/ sub) | 2026-05-31 |
| Qwen3.5: Towards Native Multimodal Agents | ✓ Link | 84.5 | Qwen3.5-9B (w/ subtitles) | 2026-02-16 |
| GLM-4.1V-Thinking: Towards Versatile Multimodal Reasoning with Scalable Reinforcement Learning | ✓ Link | 81.8 | GLM-4.6V (w/ subtitles) | 2025-07-01 |
| Qwen2.5-VL Technical Report | ✓ Link | 79.1 | Qwen2.5-VL-72B (w/ subs) | 2025-02-19 |
| HiMu: Hierarchical Multimodal Frame Selection for Long Video Question Answering | ✓ Link | 78.18 | GPT-4o + HiMu (16 frames) | 2026-03-19 |
| Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution | ✓ Link | 77.8 | Qwen2-VL-72B (w/ subtitles) | 2024-09-18 |
| InternVL3: Exploring Advanced Training and Test-Time Recipes for Open-Source Multimodal Models | ✓ Link | 75.7 | InternVL3-78B (w/ subtitles) | 2025-04-14 |
| MiniCPM-V 4.5: Cooking Efficient MLLMs via Architecture, Data, and Training Recipe | ✓ Link | 73.5 | MiniCPM-V 4.5 | 2025-09-16 |
| Ovis2.5 Technical Report | ✓ Link | 72.8 | Ovis2.5-9B (w/ subtitles) | 2025-08-15 |
| VisReflect: Latent Visual Reflection for Fine-Grained Perception in Long Visual Context | ✓ Link | 72.7 | VisReflect-7B-Video | 2026-06-29 |
| Kimi-VL Technical Report | ✓ Link | 72.6 | Kimi-VL-A3B-Instruct (w/ subs) | 2025-04-10 |
| Qwen2.5-Omni Technical Report | ✓ Link | 72.4 | Qwen2.5-Omni-7B (w/ subtitles) | 2025-03-26 |
| Aria: An Open Multimodal Native Mixture-of-Experts Model | ✓ Link | 72.1 | Aria (w/ subtitles) | 2024-10-08 |
| S3-Prune: Stability-Aware Token Budgeting for Long-Form Video-Language Models | | 71.96 | S3-prune (LLaVA-VID-7B, 25% vision-token retention) | 2026-09-08 |
| Kimi-VL Technical Report | ✓ Link | 71.9 | Kimi-VL-A3B-Thinking-2506 (w/ subs) | 2025-04-10 |
| Qwen2.5-VL Technical Report | ✓ Link | 71.6 | Qwen2.5-VL-7B | 2025-02-19 |
| Audio-Visual Flamingo: Open Audio-Visual Intelligence for Long and Complex Videos | | 71.2 | AVF-Instruct | 2026-07-17 |
| VideoLLaMA 3: Frontier Multimodal Foundation Models for Image and Video Understanding | ✓ Link | 70.3 | VideoLLaMA3 7B (w/ subs) | 2025-01-22 |
| LLaVA-OneVision: Easy Visual Task Transfer | ✓ Link | 69.5 | LLaVA-OneVision-72B (w/ subtitles) | 2024-08-06 |
| CoPE-VideoLM: Codec Primitives For Efficient Video Language Models | | 69.4 | CoPE 7B | 2026-02-13 |
| ParaVT: Taming the Tool Prior Paradox for Parallel Tool Use in Agentic Video Reinforcement Learning | ✓ Link | 69.4 | ParaVT (Qwen3-VL-8B) | 2026-05-19 |
| LiteFrame: Efficient Vision Encoders Unlock Frame Scaling in Video LLMs | ✓ Link | 68.5 | LiteFrame (w/ subtitles) | 2026-05-17 |
| FlexiVideo: Variation-Aware Temporal Dynamics Modeling for Efficient Video Understanding | ✓ Link | 67.7 | FlexiVideo-3B (extra data) | 2026-06-03 |
| Mitigating Object and Action Hallucinations in Multimodal LLMs via Self-Augmented Contrastive Alignment | | 65.6 | LLaVA-Video + SANTA | 2025-12-04 |
| AnchorPrune: Relevance-Anchored Contextual Expansion for Visual Token Pruning | ✓ Link | 60.89 | LLaVA-Video-7B (AnchorPrune, 1,024 tokens) | 2026-07-08 |
| CTEPM: Continuous-Time Event Process Memory for Long-Video Language Models | | 55.0 | CTEPM | 2026-09-08 |