| Kimi K2.5: Visual Agentic Intelligence | ✓ Link | 78.9 | Kimi K2.5 | 2026-02-02 |
| MOSS-VL Technical Report | ✓ Link | 78.0 | MOSS-VL-Instruct-0708 | 2026-08-15 |
| Rethinking Latent Visual Reasoning: Grounding Latent Reasoning in Visual Evidence | ✓ Link | 75.4 | ReaLVR (Qwen3-VL-235B-A22B) | 2026-09-28 |
| MolmoAct2: Action Reasoning Models for Real-world Deployment | ✓ Link | 72.5 | Molmo2-ER | 2026-05-04 |
| Masking Teacher and Reinforcing Student for Distilling Vision-Language Models | | 72.3 | Qwen3-VL - Masters | 2025-12-23 |
| Vero: An Open RL Recipe for General Visual Reasoning | ✓ Link | 70.9 | Vero-Qwen3.5-9B | 2026-04-06 |
| Qwen3-VL Technical Report | ✓ Link | 70.7 | Qwen3-VL-235B-A22B-Instruct | 2025-11-26 |
| Awaking Spatial Intelligence in Unified Multimodal Understanding and Generation | ✓ Link | 69.6 | JoyAI-Image-Und | 2026-05-05 |
| Penguin-VL: Exploring the Efficiency Limits of VLM with LLM-based Vision Encoders | ✓ Link | 69.1 | Penguin-VL 8B | 2026-03-06 |
| SpatialOPSD: Self-Distilling Spatial Intelligence from Verified Coding Agent Traces | | 68.9 | SpatialOPSD-9B | 2026-10-08 |
| EXAONE 4.5 Technical Report | ✓ Link | 68.8 | EXAONE 4.5 (Reasoning) | 2026-04-09 |
| OpenSpatial: A Principled Data Engine for Empowering Spatial Intelligence | ✓ Link | 68.2 | OpenSpatial-Qwen3-VL-8B | 2026-04-08 |
| STEP3-VL-10B Technical Report | ✓ Link | 67.39 | Step3-VL-10B (PaCoRe) | 2026-01-14 |
| Ovis2.5 Technical Report | ✓ Link | 67.3 | Ovis2.5-9B | 2025-08-15 |
| Scaffolding Minds: Optimizing Latent Visual Target Representations for Multimodal Reasoning | | 67.2 | Scaffolding Encoder+RL | 2026-09-29 |
| MJev-Spatial: System One Spatial Intelligence via Native Attention Matching | ✓ Link | 67.17 | MJev-Spatial | 2026-10-15 |
| Multimodal Continuous Reasoning via Asymmetric Mutual Variational Learning | | 66.91 | AMVL (7B) | 2026-07-01 |
| Context-Aware RL for Agentic and Multimodal LLMs | ✓ Link | 66.6 | ContextRL (trained from Qwen3-VL-8B) | 2026-06-15 |
| InternVL3: Exploring Advanced Training and Test-Time Recipes for Open-Source Multimodal Models | ✓ Link | 66.3 | InternVL3-78B | 2025-04-14 |
| ClaimDiff-RL: Fine-Grained Caption Reinforcement Learning through Visual Claim Comparison | ✓ Link | 66.1 | ClaimDiff-RL (actor-only) (init: Qwen3VL-32B-Instruct) | 2026-05-19 |
| Unfold The World: Factorize 4D Properties in Reinforcing Spatial Reasoning | ✓ Link | 66 | FactoSR-8B-RL | 2026-09-03 |
| GLM-4.1V-Thinking: Towards Versatile Multimodal Reasoning with Scalable Reinforcement Learning | ✓ Link | 65.5 | GLM-4.6V | 2025-07-01 |
| Toward Cognitive Supersensing in Multimodal Large Language Model | ✓ Link | 65.3 | CogSense-8B | 2026-02-02 |
| Unlocking Dense Metric Depth Estimation in VLMs | ✓ Link | 64.8 | DepthVLM-8B | 2026-05-15 |
| Qwen2.5-VL Technical Report | ✓ Link | 64.4 | Qwen2.5-VL-72B | 2025-02-19 |
| Uncovering Understanding-Generation Synergy in Native Unified Multimodal Models: From Representation, Task to System | | 64.09 | Native UMM Synergy: Dense U+G | 2026-09-01 |
| MomaGraph: State-Aware Unified Scene Graphs with Vision-Language Model for Embodied Task Planning | ✓ Link | 63.5 | MomaGraph-R1 | 2025-12-18 |
| LLaVA-OneVision-2: Towards Next-Generation Perceptual Intelligence | ✓ Link | 63.5 | LLaVA-OneVision-2-8B | 2026-05-25 |
| From Pixels to Words -- Towards Native One-Vision Models at Scale | ✓ Link | 62.8 | NEO-ov 9B | 2026-05-27 |
| Gemini 1.5: Unlocking multimodal understanding across millions of tokens of context | ✓ Link | 61.4 | Gemini 1.5 Pro | 2024-03-08 |
| Looking Back and Forth: Cross-Image Attention Calibration and Attentive Preference Learning for Multi-Image Hallucination Mitigation | | 61.33 | GLM4.1VBase+CAPL | 2026-03-07 |
| PerceptionDLM: Parallel Region Perception with Multimodal Diffusion Language Models | ✓ Link | 60.3 | PerceptionDLM-Base (8B) | 2026-06-17 |
| vLLM Semantic Router: Signal Driven Decision Routing for Mixture-of-Modality Models | ✓ Link | 59.9 | d3 | 2026-02-23 |
| VisionFoundry: Teaching VLMs Visual Perception with Synthetic Images | ✓ Link | 58.7 | MiMo-VL-7B-SFT + VisionFoundry-10K | 2026-04-10 |
| A Pocket-Sized MLLM for Ultra-Efficient Image and Video Understanding on Your Phone | | 58.7 | MiniCPM-V 4.6-Thinking (16x) | 2026-05-13 |
| Molmo2: Open Weights and Data for Vision-Language Models with Video Understanding and Grounding | ✓ Link | 57.5 | Molmo2-4B | 2026-01-15 |
| Kimi-VL Technical Report | ✓ Link | 57.3 | Kimi-VL-A3B-Instruct | 2025-04-10 |
| ProLaViT: Learning Progressive Latent Visual Thoughts in Structured Latent Space | ✓ Link | 57.23 | ProLaViT (Full) | 2026-07-03 |
| Reinforced Attention Learning | | 57.2 | RAL | 2026-02-04 |
| VisReflect: Latent Visual Reflection for Fine-Grained Perception in Long Visual Context | ✓ Link | 57.1 | VisReflect-7B-Image | 2026-06-29 |
| Phi-3 Technical Report: A Highly Capable Language Model Locally on Your Phone | ✓ Link | 57.0 | Phi-3.5-Vision | 2024-04-22 |
| Forest Before Trees: Latent Superposition for Efficient Visual Reasoning | ✓ Link | 56.92 | Laser (Qwen2.5-VL-7B-Instruct) | 2026-01-11 |
| Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction | | 56.9 | GAS (4B) | 2026-08-12 |
| EmbodiedMemory-Bench: Benchmarking Embodied Memory for Long-Horizon Embodied Tasks | ✓ Link | 56.9 | EMem-8B | 2026-09-23 |
| PerceptionLM: Open-Access Data and Models for Detailed Visual Understanding | ✓ Link | 56.0 | PLM-8B | 2025-04-17 |
| LLaVA-OneVision: Easy Visual Task Transfer | ✓ Link | 55.4 | LLaVA-OneVision-72B | 2024-08-06 |
| Unfold The World: Factorize 4D Properties in Reinforcing Spatial Understanding | | 55.4 | FactoSR-8B-RL | 2026-09-08 |
| Kwai Keye-VL 1.5 Technical Report | ✓ Link | 54.9 | Keye-VL-1.5-8B | 2025-09-01 |
| Improving Vision-language Models with Perception-centric Process Reward Models | ✓ Link | 54.49 | Qwen2.5-VL (7B) with Perceval | 2026-04-27 |
| Penguin-VL: Exploring the Efficiency Limits of VLM with LLM-based Vision Encoders | ✓ Link | 53.8 | Penguin-VL 2B | 2026-03-06 |
| PixelUMM: Encoder-Free Unified Image and Video Understanding and Generation | ✓ Link | 53.46 | PixelUMM 8B MoT | 2026-09-29 |
| Representation Forcing for Bottleneck-Free Unified Multimodal Models | | 53.0 | Pixel+RF | 2026-05-29 |
| UVU: Improving Multimodal Understanding via Vision-Language Unified Autoregressive Paradigm | | 52.8 | UVU 3B | 2026-06-03 |
| LLaVA-NeXT-Interleave: Tackling Multi-image, Video, and 3D in Large Multimodal Models | ✓ Link | 52.1 | LLaVA-NeXT-Interleave 14B | 2024-07-10 |
| Kwai Keye-VL Technical Report | ✓ Link | 52.0 | Keye-VL-8B | 2025-07-02 |
| VERDICT: Training-Free Step-Wise Verification of Multimodal Reasoning via Disagreement-Aware Consensus | | 51.32 | VERDICT | 2026-08-11 |
| ATLAS: Agentic or Latent Visual Reasoning? One Word is Enough for Both | | 51.3 | ATLAS (LA-GRPO) | 2026-05-14 |
| LFM2 Technical Report | ✓ Link | 51.03 | LFM2-VL-3B | 2025-11-28 |
| Introducing LFM2.5: The Next Generation of On-Device AI | | 48.82 | LFM2.5-VL-1.6B | 2026-01-05 |
| InfiniteVL: Synergizing Linear and Sparse Attention for Highly-Efficient, Unlimited-Input Vision-Language Models | ✓ Link | 47.6 | InfiniteVL-Base | 2025-12-09 |
| ZAYA1-VL-8B Technical Report | ✓ Link | 45.9 | ZAYA1-VL-8B | 2026-05-08 |
| VideoLLaMA 3: Frontier Multimodal Foundation Models for Image and Video Understanding | ✓ Link | 44.2 | VideoLLaMA3 2B | 2025-01-22 |
| LFM2.5-VL-450M: Structured Visual Intelligence, Edge to Cloud | | 43.92 | LFM2.5-VL-450M | 2026-04-08 |
| Boosting Visual Instruction Tuning with Self-Supervised Guidance | ✓ Link | 41.8 | LLaVA-1.5-Qwen2.5-7B + V-GIFT | 2026-04-14 |
| Rethinking Projector Training in Multimodal LLMs | ✓ Link | 41.5 | PORTAL (Qwen 2.5 7B + CLIP) | 2026-10-13 |
| Omni-R1: Towards the Unified Generative Paradigm for Multimodal Reasoning | ✓ Link | 35.61 | Omni-R1-M | 2026-01-14 |