| Qwen3.5: Towards Native Multimodal Agents | ✓ Link | 90.3 | Qwen3.5-397B-A17B | 2026-02-16 |
| Kimi K2.5: Visual Agentic Intelligence | ✓ Link | 90.1 | Kimi K2.5 | 2026-02-02 |
| HopChain: Multi-Hop Data Synthesis for Generalizable Vision-Language Reasoning | | 89 | Qwen3.5-397B-A17B (RLVR w/ Multi-Hop) | 2026-03-17 |
| Qwen3.5: Towards Native Multimodal Agents | ✓ Link | 87.8 | Qwen3.5-27B | 2026-02-16 |
| Qwen3.5: Towards Native Multimodal Agents | ✓ Link | 87.4 | Qwen3.5-122B-A10B | 2026-02-16 |
| Qwen3.6 | ✓ Link | 87.4 | Qwen3.6-27B | 2026-04-21 |
| Introducing OpenAI o3 and o4-mini | ✓ Link | 86.8 | OpenAI o3 | 2025-04-16 |
| Scalable Visual Pretraining for Language Intelligence | | 86.70 | Qwen3.5-35B-A3B + VP | 2026-07-10 |
| Beyond Correctness: Benchmarking and Aligning Response Behaviors in Hybrid-Thinking MLLMs | | 86.61 | Qwen3-VL-8B + PatternRL | 2026-08-17 |
| Qwen3.6 | ✓ Link | 86.4 | Qwen3.6-35B-A3B | 2026-04-21 |
| Qwen3.5: Towards Native Multimodal Agents | ✓ Link | 86.2 | Qwen3.5-35B-A3B | 2026-02-16 |
| Qwen3.5-Omni Technical Report | | 86.1 | Qwen3.5-Omni-Plus | 2026-04-21 |
| SenseNova-U1.5: Towards Native Unified Visual Intelligence | ✓ Link | 85.85 | SenseNova-U1.5 8B-Think | 2026-09-10 |
| Qwen3-VL Technical Report | ✓ Link | 85.8 | Qwen3-VL-235B-A22B-Thinking | 2025-11-26 |
| Thinking with Comics: Enhancing Multimodal Reasoning through Structured Visual Storytelling | ✓ Link | 85.8 | TwC - Img & Txt (Generate-then-Reason) | 2026-02-02 |
| Qwen3.5: Towards Native Multimodal Agents | ✓ Link | 85.7 | Qwen3.5-9B | 2026-02-16 |
| MAXS: Meta-Adaptive Exploration with LLM Agents | ✓ Link | 85.5 | MAXS (MiMo-VL-7B-SFT) | 2026-01-14 |
| STEP3-VL-10B Technical Report | ✓ Link | 85.5 | Step3-VL-10B (PaCoRe) | 2026-01-14 |
| SenseNova-U1: Unifying Multimodal Understanding and Generation with NEO-unify Architecture | ✓ Link | 85.30 | SenseNova-U1-A3B-MoT | 2026-05-12 |
| GLM-4.1V-Thinking: Towards Versatile Multimodal Reasoning with Scalable Reinforcement Learning | ✓ Link | 85.2 | GLM-4.6V | 2025-07-01 |
| Masking Teacher and Reinforcing Student for Distilling Vision-Language Models | | 85.0 | InternVL3.5 - Masters | 2025-12-23 |
| EXAONE 4.5 Technical Report | ✓ Link | 85.0 | EXAONE 4.5 (Reasoning) | 2026-04-09 |
| Group Entropy-Controlled Policy Optimization | | 84.9 | GEPO (Qwen3.5-9B) | 2026-07-18 |
| ERNIE 5.0 Technical Report | | 84.8 | ERNIE 5.0 | 2026-02-04 |
| Evidence-RL: Towards Evidence-intensive Visual Reasoning | | 83.75 | Evidence-RL (Qwen3.5-9B) | 2026-08-08 |
| Ovis2.5 Technical Report | ✓ Link | 83.4 | Ovis2.5-9B | 2025-08-15 |
| LongCat-Next: Lexicalizing Modalities as Discrete Tokens | ✓ Link | 83.1 | LongCat-Next (A3B, 68.5B) | 2026-03-29 |
| Kwai Keye-VL-2.0 Technical Report | ✓ Link | 82.9 | Keye-VL-2.0-30B-A3B | 2026-06-09 |
| Nemotron 3 Nano Omni: Efficient and Open Multimodal Intelligence | ✓ Link | 82.8 | Nemotron 3 Nano Omni (Reasoning) | 2026-04-27 |
| Self-Supervised Visual On-Policy Distillation | | 81.5 | S^2VOPD (Qwen3.5-4B, Vision-OPD-6K) (extra data) | 2026-08-14 |
| V-Rubrics: Visual Faithfulness via Rubric-Based Reinforcement Learning | ✓ Link | 81.3 | V-Rubrics (GRPO w/ rubrics) (extra data) | 2026-08-26 |
| Kwai Keye-VL 1.5 Technical Report | ✓ Link | 81.2 | Keye-VL-1.5-8B | 2025-09-01 |
| Qwen3-Omni Technical Report | ✓ Link | 81.2 | Qwen3-Omni-Flash-Thinking | 2025-09-22 |
| AnE: Pushing the Reasoning Frontier of Multimodal LLMs via Anchor Evolution | ✓ Link | 81.2 | AnE-3rd | 2026-05-25 |
| Kwai Keye-VL Technical Report | ✓ Link | 80.7 | Keye-VL-8B | 2025-07-02 |
| Command A+ | | 80.6 | Command A+ | 2026-05-20 |
| See Less, See Right: Bi-directional Perceptual Shaping For Multimodal Reasoning | ✓ Link | 80.4 | BiPS (Qwen3-VL-8B-Thinking) | 2025-12-26 |
| EvoCUA: Evolving Computer Use Agents via Learning from Scalable Synthetic Experience | ✓ Link | 80.40 | EvoCUA-32B | 2026-01-22 |
| Kimi-VL Technical Report | ✓ Link | 80.1 | Kimi-VL-A3B-Thinking-2506 | 2025-04-10 |
| MiniCPM-o 4.5: Towards Real-Time Full-Duplex Omni-Modal Interaction | ✓ Link | 80.1 | MiniCPM-o 4.5-Instruct | 2026-04-30 |
| MiniCPM-V 4.5: Cooking Efficient MLLMs via Architecture, Data, and Training Recipe | ✓ Link | 79.9 | MiniCPM-V 4.5 | 2025-09-16 |
| Context-Aware RL for Agentic and Multimodal LLMs | ✓ Link | 79.8 | ContextRL (trained from Qwen3-VL-8B) | 2026-06-15 |
| OpenVLThinkerV2: A Generalist Multimodal Reasoning Model for Multi-domain Visual Tasks | ✓ Link | 79.5 | OpenVLThinkerV2 (Ours) | 2026-04-09 |
| Near-Future Policy Optimization | | 79.2 | AutoNPO | 2026-04-22 |
| Learning Consistency in Reward Modeling for Multi-Modal Reasoning | | 79.2 | CRM-RL-8B (Ours, soft-reward) | 2026-09-08 |
| Can We Trust the Teacher? Decoupled Credit Direction-Magnitude for Self-Distillation | | 79.2 | DCSD (Qwen3-VL-8B-Instruct) | 2026-09-28 |
| InternVL3: Exploring Advanced Training and Test-Time Recipes for Open-Source Multimodal Models | ✓ Link | 79.0 | InternVL3-78B | 2025-04-14 |
| ERNIE 4.5 Technical Report | ✓ Link | 78.9 | ERNIE-4.5-VL-424B-A47B | 2025-06-30 |
| Self-Distilled RLVR | | 78.1 | RLSD (Qwen3-VL-8B-Instruct) | 2026-04-03 |
| Act Wisely: Cultivating Meta-Cognitive Tool Use in Agentic Multimodal Models | ✓ Link | 78.0 | Metis | 2026-04-09 |
| OneThinker: All-in-one Reasoning Model for Image and Video | ✓ Link | 77.6 | OneThinker-8B | 2025-12-02 |
| Insight-V++: Towards Advanced Long-Chain Visual Reasoning with Multimodal Large Language Models | | 77.6 | Insight-V++ | 2026-03-18 |
| Penguin-VL: Exploring the Efficiency Limits of VLM with LLM-based Vision Encoders | ✓ Link | 77.2 | Penguin-VL 8B | 2026-03-06 |
| LLaVA-UHD v4: What Makes Efficient Visual Encoding in MLLMs? | ✓ Link | 76.9 | LLaVA-UHD v4 | 2026-05-09 |
| Dynamic Important Example Mining for Reinforcement Finetuning | ✓ Link | 76.9 | Qwen2.5-VL-32B + DIEM | 2026-08-29 |
| CogFlow: Bridging Perception and Reasoning through Knowledge Internalization for Visual Mathematical Problem Solving | ✓ Link | 76.8 | CogFlow (7B) | 2026-01-05 |
| Co-Evolving Policy Distillation | | 76.8 | CoPD | 2026-04-29 |
| Thinking in Uncertainty: Mitigating Hallucinations in MLRMs with Latent Entropy-Aware Decoding | ✓ Link | 76.3 | VL-Cogito-7B with LEAD | 2026-03-09 |
| CodePercept: Code-Grounded Visual STEM Perception for MLLMs | ✓ Link | 75.9 | CodePercept (32B, S1) | 2026-03-11 |
| From Seeing to Thinking: Decoupling Perception and Reasoning Improves Post-Training of Vision-Language Models | ✓ Link | 75.9 | Qwen3-VL-8B (staged training) | 2026-05-19 |
| Beyond Layer-Wise Merging: Chain-of-Merging for Vision-Language Models | | 75.9 | CoM (Qwen2.5-VL-72B + Qwen2.5-Math-72B) | 2026-06-03 |
| Beyond Where to Look: Trajectory-Guided Reinforcement Learning for Multimodal RLVR | | 75.0 | TGRL-DAPO (Qwen2.5-VL-7B, 39K) | 2026-03-27 |
| Unlocking Token Rewards via Training-Free Reward Attribution | | 75.0 | Qwen2.5-VL-7B-Instruct + GRPO (P2T token reward) (extra data) | 2026-06-03 |
| Qwen2.5-VL Technical Report | ✓ Link | 74.8 | Qwen2.5-VL-72B | 2025-02-19 |
| From Narrow to Panoramic Vision: Attention-Guided Cold-Start Reshapes Multimodal Reasoning | ✓ Link | 74.7 | AVAR-Thinker | 2026-03-04 |
| AdaTooler-V: Adaptive Tool-Use for Images and Videos | ✓ Link | 74.5 | AdaTooler-V-7B | 2025-12-18 |
| Step-GUI Technical Report | ✓ Link | 74.4 | Step-GUI-8B | 2025-12-17 |
| Length-Unbiased Sequence Policy Optimization: Revealing and Controlling Response Length Variation in RLVR | ✓ Link | 74.4 | Qwen2.5-VL-7B-Instruct (LUSPO) | 2026-02-05 |
| Awaking Spatial Intelligence in Unified Multimodal Understanding and Generation | ✓ Link | 74.4 | JoyAI-Image-Und | 2026-05-05 |
| SPOT-E: Test-Time Entropy Shaping with Visual Spotlights for Frozen VLMs | ✓ Link | 74.4 | SPOT-E (InternVL3-8B) | 2026-06-18 |
| Uni-Edit: Intelligent Editing Is A General Task For Unified Model Tuning | ✓ Link | 74.3 | BAGEL-Uni-Edit (stage 2) | 2026-05-20 |
| Uncovering Understanding-Generation Synergy in Native Unified Multimodal Models: From Representation, Task to System | | 74.3 | Native UMM Synergy: Und+Gen geometry | 2026-09-01 |
| PointThinker: Point-Incentivized Parallel Thinking for Multimodal Large Language Model | | 74.2 | PointThinker-7B (extra data) | 2026-06-03 |
| Before Thinking, Learn to Decide: Proactive Routing for Efficient Visual Reasoning | | 74.2 | PRP (JRL, 7B) | 2026-06-29 |
| Introducing OpenAI o1 | | 73.9 | OpenAI o1 | 2024-09-12 |
| The Llama 4 herd: The beginning of a new era of natively multimodal AI innovation | ✓ Link | 73.7 | Llama 4 Maverick | 2025-04-05 |
| VideoAuto-R1: Video Auto Reasoning via Thinking Once, Answering Twice | ✓ Link | 73.7 | VideoAuto-R1 (Qwen2.5-VL-7B) | 2026-01-08 |
| Introducing Command A Vision: Multimodal AI built for business | | 73.5 | Command A Vision | 2025-07-31 |
| Emerging Properties in Unified Multimodal Pretraining | ✓ Link | 73.1 | BAGEL (7B MoT) | 2025-05-20 |
| LASER: A Corrective Lens for LVLMs via Visual Attention Preservation and Sink Suppression | ✓ Link | 72.9 | LASER | 2026-07-02 |
| CPPO: Contrastive Perception for Vision Language Policy Optimization | ✓ Link | 72.2 | CPPO (Qwen2.5-VL-7B) | 2026-01-01 |
| UniPath: Adaptive Coordination of Understanding and Generation for Unified Multimodal Reasoning | | 72.2 | UniPath (BAGEL backbone) | 2026-05-12 |
| Improving Vision-language Models with Perception-centric Process Reward Models | ✓ Link | 72.0 | Qwen2.5-VL (7B) with Perceval | 2026-04-27 |
| CodeV: Code with Images for Faithful Visual Reasoning via Tool-Aware Policy Optimization | ✓ Link | 71.8 | CodeV-7B-RL | 2025-11-24 |
| Puzzle Curriculum GRPO for Vision-Centric Reasoning | ✓ Link | 71.7 | Qwen2.5-VL-7B (Rotation) | 2025-12-16 |
| SVR-R1: Bootstrapping Multi-modal Reasoning with Self-verification in Reinforcement Learning | | 71.6 | SVR-R1 (7B ThinkLite) | 2026-07-13 |
| Infinity-Parser2-Pro | | 71.4 | Infinity-Parser2-Pro | 2026-05-11 |
| Generation as Auxiliary Supervision: Enhancing Visual Understanding at Zero Inference Overhead via Decoupled Embedding Prediction | | 71.2 | GAS (2B after pretraining) | 2026-08-12 |
| Co-RL: Unsupervised Reasoning Emerges from Diverse Cohort in Multi-agent RL | ✓ Link | 71.0 | Co-RL (Qwen2.5-VL-7B, Different family) | 2026-08-19 |
| The Llama 4 herd: The beginning of a new era of natively multimodal AI innovation | ✓ Link | 70.7 | Llama 4 Scout | 2025-04-05 |
| Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution | ✓ Link | 70.5 | Qwen2-VL-72B | 2024-09-18 |
| LLaDA-UI: Bringing Block-wise Diffusion to Vision-Language GUI Agents | ✓ Link | 70.5 | LLaDA-UI (pre-training) | 2026-09-09 |
| Intern Lumina U2: A Multi-Codebook Diffusion Large Language Model for Omni-Visual Understanding and Image Generation | ✓ Link | 70.4 | Intern Lumina U2 (NVIDIA) | 2026-10-14 |
| ARM-Thinker: Reinforcing Multimodal Generative Reward Models with Agentic Tool Use and Visual Reasoning | ✓ Link | 70.2 | ARM-Thinker-7B | 2025-12-04 |
| Language-guided Frequency Modulation for Large Vision-Language Models | | 69.8 | LFM (w/ Qwen2.5-VL) | 2026-06-03 |
| Kimi-VL Technical Report | ✓ Link | 68.7 | Kimi-VL-A3B-Instruct | 2025-04-10 |
| MiniMax-01: Scaling Foundation Models with Lightning Attention | ✓ Link | 68.6 | MiniMax-VL-01 | 2025-01-14 |
| Decouple to Generalize: Context-First Self-Evolving Learning for Data-Scarce Vision-Language Reasoning | ✓ Link | 68.3 | DoGe-7B (Iter3) | 2025-12-07 |
| Qwen2.5-VL Technical Report | ✓ Link | 68.2 | Qwen2.5-VL-7B | 2025-02-19 |
| LLaDA2.0-Uni: Unifying Multimodal Understanding and Generation with Diffusion Large Language Model | ✓ Link | 68.1 | LLaDA2.0-Uni (16B) | 2026-04-22 |
| Qwen2.5-Omni Technical Report | ✓ Link | 67.9 | Qwen2.5-Omni-7B | 2025-03-26 |
| Learning from Primitive: Probing Visual Reasoning of LVLMs via Counting | | 67.9 | Qwen3-VL-8B (Learning from Primitive) (extra data) | 2026-09-08 |
| Introducing Claude 3.5 Sonnet | | 67.7 | Claude 3.5 Sonnet | 2024-06-20 |
| Gemma 3 Technical Report | ✓ Link | 67.6 | Gemma 3 (27B, IT) | 2025-03-25 |
| LLaVA-OneVision: Easy Visual Task Transfer | ✓ Link | 67.5 | LLaVA-OneVision-72B | 2024-08-06 |
| VideoLLaMA 3: Frontier Multimodal Foundation Models for Image and Video Understanding | ✓ Link | 67.1 | VideoLLaMA3 7B | 2025-01-22 |
| Reasoning Within the Mind: Dynamic Multimodal Interleaving in Latent Space | ✓ Link | 66.9 | Qwen3-VL-8B + DMLR | 2025-12-14 |
| HyperVL: An Efficient and Dynamic Multimodal Large Language Model for Edge Devices | | 66.2 | HyperVL | 2025-12-16 |
| Aria: An Open Multimodal Native Mixture-of-Experts Model | ✓ Link | 66.1 | Aria | 2024-10-08 |
| LLaDA-o: An Effective and Length-Adaptive Omni Diffusion Model | ✓ Link | 66.1 | LLaDA-o | 2026-03-01 |
| AdaptVision: Efficient Vision-Language Models via Adaptive Visual Acquisition | ✓ Link | 65.9 | AdaptVision (Qwen2.5-VL-7B-Instruct) | 2025-12-03 |
| PerceptionDLM: Parallel Region Perception with Multimodal Diffusion Language Models | ✓ Link | 65.5 | PerceptionDLM-Base (8B) | 2026-06-17 |
| InfiniteVL: Synergizing Linear and Sparse Attention for Highly-Efficient, Unlimited-Input Vision-Language Models | ✓ Link | 65.4 | InfiniteVL-Base | 2025-12-09 |
| Label-free steering: Compressing test-time reinforcement learning into bias-only subspaces | ✓ Link | 65.4 | Qwen2.5-VL-7B-Instruct + Label-Free Bias-Only TTRL | 2026-09-28 |
| DiffusionVL: Translating Any Autoregressive Models into Diffusion Vision Language Models | ✓ Link | 64.3 | DiffusionVL | 2025-12-17 |
| NaVLM-PVC: Progressive Visual Compression for Efficient Native-Resolution Encoding in MLLMs | | 64.2 | NaVLM-PVC | 2026-09-08 |
| ZAYA1-VL-8B Technical Report | ✓ Link | 64.0 | ZAYA1-VL-8B | 2026-05-08 |
| Process Rewards with Learned Reliability | ✓ Link | 64.0 | ACA with BetaPRM (Qwen2.5-VL-7B) | 2026-05-15 |
| Gemini 1.5: Unlocking multimodal understanding across millions of tokens of context | ✓ Link | 63.9 | Gemini 1.5 Pro | 2024-03-08 |
| Hello GPT-4o | | 63.8 | GPT-4o | 2024-05-13 |
| Dream-VL & Dream-VLA: Open Vision-Language and Vision-Language-Action Models with Diffusion Language Model Backbone | ✓ Link | 63.1 | Dream-VL | 2025-12-27 |
| VisionFoundry: Teaching VLMs Visual Perception with Synthetic Images | ✓ Link | 62.9 | Qwen2.5-VL-3B-Instruct + VisionFoundry-10K | 2026-04-10 |
| Phi-4-Mini Technical Report: Compact yet Powerful Multimodal Language Models via Mixture-of-LoRAs | ✓ Link | 62.4 | Phi-4-Multimodal | 2025-02-24 |
| Wake up for Touch! Mask-isolated Tactile Alignment Learning in MLLMs | ✓ Link | 62.3 | Splash (3B) | 2026-07-01 |
| LFM2 Technical Report | ✓ Link | 62.20 | LFM2-VL-3B | 2025-11-28 |
| Penguin-VL: Exploring the Efficiency Limits of VLM with LLM-based Vision Encoders | ✓ Link | 61.3 | Penguin-VL 2B | 2026-03-06 |
| VideoLLaMA 3: Frontier Multimodal Foundation Models for Image and Video Understanding | ✓ Link | 59.2 | VideoLLaMA3 2B | 2025-01-22 |
| Training Data Efficiency in Multimodal Process Reward Models | ✓ Link | 59.0 | InternVL2.5-8B with MPRM (BIS-25%) | 2026-02-04 |
| Molmo2: Open Weights and Data for Vision-Language Models with Video Understanding and Grounding | ✓ Link | 58.9 | Molmo2-8B | 2026-01-15 |
| Molmo and PixMo: Open Weights and Open Data for State-of-the-Art Vision-Language Models | ✓ Link | 58.6 | Molmo 72B | 2024-09-25 |
| Building and better understanding vision-language models: insights and future directions | ✓ Link | 58.4 | Idefics3-8B | 2024-08-22 |
| Pixtral 12B | ✓ Link | 58.3 | Pixtral 12B | 2024-10-09 |
| Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution | ✓ Link | 58.2 | Qwen2-VL-7B | 2024-09-18 |
| Llama 3.2: Revolutionizing edge AI and vision with open, customizable models | ✓ Link | 57.3 | Llama 3.2 90B Vision Instruct | 2024-09-25 |
| Sparse-LaViDa: Sparse Multimodal Discrete Diffusion Language Models | | 56.7 | Sparse-LaViDa | 2025-12-16 |
| MiniCPM-V: A GPT-4V Level MLLM on Your Phone | ✓ Link | 54.3 | MiniCPM-Llama3-V 2.5 | 2024-08-03 |
| LoMo: Local Modality Substitution for Deeper Vision-Language Fusion | ✓ Link | 53.9 | LoMo (LLaVA-OneVision-1.5-8B) | 2026-05-28 |
| AutoNeural: Co-Designing Vision-Language Models for NPU Inference | | 53.1 | AutoNeural | 2025-12-02 |
| Gemini: A Family of Highly Capable Multimodal Models | ✓ Link | 53.0 | Gemini Ultra | 2023-12-19 |
| What matters when building vision-language models? | ✓ Link | 51.6 | Idefics2 (w/o image split) | 2024-05-03 |
| SmolVLM: Redefining small and efficient multimodal models | ✓ Link | 51.5 | SmolVLM 2.2B | 2025-04-07 |
| Cheers: Decoupling Patch Details from Semantic Representations Enables Unified Multimodal Comprehension and Generation | ✓ Link | 50.5 | Cheers (1.5B) | 2026-03-13 |
| Learning Adaptive Reasoning Paths for Efficient Visual Reasoning | ✓ Link | 50.2 | AVR-8B | 2026-04-16 |
| LLaVA-NeXT: Improved reasoning, OCR, and world knowledge | ✓ Link | 46.5 | LLaVA-NeXT-34B | 2024-01-30 |
| Phi-3 Technical Report: A Highly Capable Language Model Locally on Your Phone | ✓ Link | 43.9 | Phi-3.5-Vision | 2024-04-22 |
| Ministral 3 | | 43.6 | Ministral 3 (14B) | 2026-01-13 |
| Ovis: Structural Embedding Alignment for Multimodal Large Language Model | ✓ Link | 43.4 | Ovis-Qwen1.5-14B | 2024-05-31 |
| Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution | ✓ Link | 43.0 | Qwen2-VL-2B | 2024-09-18 |
| Falcon2-11B Technical Report | | 37.2 | Falcon2-11B VLM | 2024-07-20 |
| DeepSeek-VL: Towards Real-World Vision-Language Understanding | ✓ Link | 36.1 | DeepSeek-VL 7B Chat | 2024-03-08 |
| Rethinking Projector Training in Multimodal LLMs | ✓ Link | 31.9 | PORTAL (Qwen 2.5 7B + CLIP) | 2026-10-13 |