mlsanityData to 7 Oct 2026

Qwen3-VL Technical Report

arXiv 2511.21631 ↗First posted 26 Nov 2025cs.CV cs.AIcited by 2,534

On the map: Multimodal LLMs / vision-language models → MLLM training and visual instruction tuning

also in 1 more
Abstract (arXiv)

We introduce Qwen3-VL, the most capable vision-language model in the Qwen series to date, achieving superior performance across a broad range of multimodal benchmarks. It natively supports interleaved contexts of up to 256K tokens, seamlessly integrating text, images, and video. The model family includes both dense (2B/4B/8B/32B) and mixture-of-experts (30B-A3B/235B-A22B) variants to accommodate diverse latency-quality trade-offs. Qwen3-VL delivers three core pillars: (i) markedly stronger pure-text understanding, surpassing comparable text-only backbones in several cases; (ii) robust long-context comprehension with a native 256K-token window for both text and interleaved multimodal inputs, enabling faithful retention, retrieval, and cross-referencing across long documents and videos; and (iii) advanced multimodal reasoning across single-image, multi-image, and video tasks, demonstrating leading performance on comprehensive evaluations such as MMMU and visual-math benchmarks (e.g., MathVista and MathVision). Architecturally, we introduce three key upgrades: (i) an enhanced interleaved-MRoPE for stronger spatial-temporal modeling across images and video; (ii) DeepStack integration, which effectively leverages multi-level ViT features to tighten vision-language alignment; and (iii) text-based time alignment for video, evolving from T-RoPE to explicit textual timestamp alignment for more precise temporal grounding. Under comparable token budgets and latency constraints, Qwen3-VL achieves superior performance in both dense and Mixture-of-Experts (MoE) architectures. We envision Qwen3-VL serving as a foundational engine for image-grounded reasoning, agentic decision-making, and multimodal code intelligence in real-world workflows.

Key facts. Cited by 2,534 papers in the map (data through 7 Oct 2026). Placed in MLLM training and visual instruction tuning, one of 16 directions in Multimodal LLMs / vision-language models. Builds on 84 of its 102 references that are in the map. First posted on arXiv 26 Nov 2025.

Builds on

84 of 102 references.

Closest ancestors · 10

Same direction, newest first

  1. 2025SimpleVQA: Multimodal Factuality Evaluation for Multimodal Large Language Models18 Feb
  2. 2025ZeroBench: An Impossible Visual Benchmark for Contemporary Large Multimodal Models13 Feb
  3. 2024Pixtral 12B9 Oct
  4. 2024Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution18 Sep
  5. 2024Divide, Conquer and Combine: A Training-Free Framework for High-Resolution Image Perception in Multimodal Large Language Models28 Aug
  6. 2024LogicVista: Multimodal LLM Logical Reasoning Benchmark in Visual Contexts6 Jul
  7. 2024MIA-Bench: Towards Better Instruction Following Evaluation of Multimodal LLMs1 Jul
  8. 2024MuirBench: A Comprehensive Benchmark for Robust Multi-image Understanding13 Jun
  9. 2024BLINK: Multimodal Large Language Models Can See but Not Perceive18 Apr
  10. 2023OBELICS: An Open Web-Scale Filtered Dataset of Interleaved Image-Text Documents21 Jun

Other references · 74

  1. 2014Microsoft COCO: Common Objects in Contextcited by 17,498
  2. 2025Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilitiescited by 3,655
  3. 2023Grounding DINO: Marrying DINO with Grounded Pre-Training for Open-Set Object Detectioncited by 3,274
  4. 2023GPQA: A Graduate-Level Google-Proof Q&A Benchmarkcited by 2,748
  5. 2017Billion-scale similarity search with GPUscited by 2,477
  6. 2023MMMU: A Massive Multi-discipline Multimodal Understanding and Reasoning Benchmark for Expert AGIcited by 2,044
  7. 2024LiveCodeBench: Holistic and Contamination Free Evaluation of Large Language Models for Codecited by 1,906
  8. 2024MMLU-Pro: A More Robust and Challenging Multi-Task Language Understanding Benchmarkcited by 1,758
  9. 2023MathVista: Evaluating Mathematical Reasoning of Foundation Models in Visual Contextscited by 1,582
  10. 2022ChartQA: A Benchmark for Question Answering about Charts with Visual and Logical Reasoningcited by 1,449
  11. 2024Video-MME: The First-Ever Comprehensive Evaluation Benchmark of Multi-modal LLMs in Video Analysiscited by 1,375
  12. 2025Search-R1: Training LLMs to Reason and Leverage Search Engines with Reinforcement Learningcited by 1,371
38 more
  1. 2020DocVQA: A Dataset for VQA on Document Imagescited by 1,288
  2. 2021Grounded Language-Image Pre-trainingcited by 1,147
  3. 2023Instruction-Following Evaluation for Large Language Modelscited by 1,143
  4. 2024OSWorld: Benchmarking Multimodal Agents for Open-Ended Tasks in Real Computer Environmentscited by 1,134
  5. 2025SigLIP 2: Multilingual Vision-Language Encoders with Improved Semantic Understanding, Localization, and Dense Featurescited by 1,127
  6. 2023MVBench: A Comprehensive Multi-modal Video Understanding Benchmarkcited by 1,086
  7. 2015Generation and Comprehension of Unambiguous Object Descriptionscited by 897
  8. 2024Are We on the Right Way for Evaluating Large Vision-Language Models?cited by 893
  9. 2016A Diagram Is Worth A Dozen Imagescited by 793
  10. 2020Hypersim: A Photorealistic Synthetic Dataset for Holistic Indoor Scene Understandingcited by 593
  11. 2017TALL: Temporal Activity Localization via Language Querycited by 581
  12. 2024Thinking in Space: How Multimodal Large Language Models See, Remember, and Recall Spacescited by 580
  13. 2024MathVerse: Does Your Multi-modal LLM Truly See the Diagrams in Visual Math Problems?cited by 573
  14. 2024SeeClick: Harnessing GUI Grounding for Advanced Visual GUI Agentscited by 553
  15. 2024Measuring Multimodal Mathematical Reasoning with MATH-Vision Datasetcited by 516
  16. 2024MMMU-Pro: A More Robust Multi-discipline Multimodal Understanding Benchmarkcited by 494
  17. 2024From Crowdsourced Data to High-Quality Benchmarks: Arena-Hard and BenchBuilder Pipelinecited by 484
  18. 2023OCRBench: On the Hidden Mystery of OCR in Large Multimodal Modelscited by 466
  19. 2023V*: Guided Visual Search as a Core Mechanism in Multimodal LLMscited by 459
  20. 2024AndroidWorld: A Dynamic Benchmarking Environment for Autonomous Agentscited by 450
  21. 2021ARKitScenes: A Diverse Real-World Dataset For 3D Indoor Scene Understanding Using Mobile RGB-D Datacited by 442
  22. 2021InfographicVQAcited by 430
  23. 2024LVBench: An Extreme Long Video Understanding Benchmarkcited by 402
  24. 2025DeepEyes: Incentivizing "Thinking with Images" via Reinforcement Learningcited by 398
  25. 2024We-Math: Does Your Large Multimodal Model Achieve Human-like Mathematical Reasoning?cited by 281
  26. 2025ScreenSpot-Pro: GUI Grounding for Professional High-Resolution Computer Usecited by 246
  27. 2024CharXiv: Charting Gaps in Realistic Chart Understanding in Multimodal LLMscited by 229
  28. 2025Video-MMMU: Evaluating Knowledge Acquisition from Multi-Discipline Professional Videoscited by 225
  29. 2023Multimodal C4: An Open, Billion-scale Corpus of Images Interleaved with Textcited by 208
  30. 2023Teaching CLIP to Count to Tencited by 194
  31. 2025Mobile-Agent-v3: Fundamental Agents for GUI Automationcited by 183
  32. 2024MMLongBench-Doc: Benchmarking Long-context Document Understanding with Visualizationscited by 174
  33. 2024OCRBench v2: An Improved Benchmark for Evaluating Large Multimodal Models on Visual Text Localization and Reasoningcited by 165
  34. 2022Omni3D: A Large Benchmark and Model for 3D Object Detection in the Wildcited by 159
  35. 2025OpenCUA: Open Foundations for Computer-Use Agentscited by 151
  36. 2024Design2Code: Benchmarking Multimodal Code Generation for Automated Front-End Engineeringcited by 146
  37. 2024DynaMath: A Dynamic Visual Benchmark for Evaluating Mathematical Reasoning Robustness of Vision Language Modelscited by 143
  38. 2024RoboSpatial: Teaching Spatial Understanding to 2D and 3D Vision-Language Models for Roboticscited by 133

50 most cited of 74 All 74 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 73; 2026: 2461.

2,534 papers.

Sort:

By area:

  1. 2026Kimi K2.5: Visual Agentic Intelligencecited by 449
  2. 2026Qwen3-VL-Embedding and Qwen3-VL-Reranker: A Unified Framework for State-of-the-Art Multimodal Retrieval and Rankingcited by 211
  3. 2026Advancing Open-source World Modelscited by 144
  4. 2026Cosmos 3: Omnimodal World Models for Physical AIcited by 137
  5. 2026Molmo2: Open Weights and Data for Vision-Language Models with Video Understanding and Groundingcited by 132
  6. 2026StarVLA: A Lego-like Codebase for Vision-Language-Action Model Developingcited by 103
  7. 2026ABot-M0: VLA Foundation Model for Robotic Manipulation with Action Manifold Learningcited by 94
  8. 2025MAI-UI Technical Report: Real-World Centric Foundation GUI Agentscited by 88
  9. 2026OpenClaw-RL: Train Any Agent Simply by Talkingcited by 83
  10. 2026Mobile-Agent-v3.5: Multi-platform Fundamental GUI Agentscited by 72
  11. 2026DiT4DiT: Jointly Modeling Video Dynamics and Actions for Generalizable Robot Controlcited by 71
  12. 2026XSkill: Continual Learning from Experience and Skills in Multimodal Agentscited by 71
38 more
  1. 2026DeepSeek-OCR 2: Visual Causal Flowcited by 68
  2. 2026WorldArena: A Unified Benchmark for Evaluating Perception and Functional Utility of Embodied World Modelscited by 68
  3. 2026Motubrain: An Advanced World Action Model for Robot Controlcited by 54
  4. 2025OneThinker: All-in-one Reasoning Model for Image and Videocited by 52
  5. 2026EvoCUA: Evolving Computer Use Agents via Learning from Scalable Synthetic Experiencecited by 51
  6. 2026Robometer: Scaling General-Purpose Robotic Reward Models via Trajectory Comparisonscited by 51
  7. 2025WorldMM: Dynamic Multimodal Memory Agent for Long Video Reasoningcited by 50
  8. 2026Zooming without Zooming: Region-to-Image Distillation for Fine-Grained Multimodal Perceptioncited by 47
  9. 2026Qwen-RobotManip Technical Report: Alignment Unlocks Scale for Robotic Manipulation Foundation Modelscited by 47
  10. 2025TimeLens: Rethinking Video Temporal Grounding with Multimodal LLMscited by 46
  11. 2026WoVR: World Models as Reliable Simulators for Post-Training VLA Policies with RLcited by 46
  12. 2026Video-MME-v2: Towards the Next Stage in Benchmarks for Comprehensive Video Understandingcited by 46
  13. 2026MolmoAct2: Action Reasoning Models for Real-world Deploymentcited by 45
  14. 2026Uni-OPD: Unifying On-Policy Distillation with a Dual-Perspective Recipecited by 43
  15. 2026Xiaomi-Robotics-0: An Open-Sourced Vision-Language-Action Model with Real-Time Executioncited by 42
  16. 2026Qwen-VLA: Unifying Vision-Language-Action Modeling across Tasks, Environments, and Robot Embodimentscited by 42
  17. 2026BabyVision: Visual Reasoning Beyond Languagecited by 39
  18. 2026UI-Venus-1.5 Technical Reportcited by 39
  19. 2025Step-GUI Technical Reportcited by 38
  20. 2026VLM4VLA: Revisiting Vision-Language-Models in Vision-Language-Action Modelscited by 38
  21. 2026Harness VLA: Steering Frozen VLAs into Reliable Manipulation Primitives via Memory-Guided Agentscited by 37
  22. 2026Lyra 2.0: Explorable Generative 3D Worldscited by 36
  23. 2026Kiwi-Edit: Versatile Video Editing via Instruction and Reference Guidancecited by 34
  24. 2026WBench: A Comprehensive Multi-turn Benchmark for Interactive Video World Model Evaluationcited by 34
  25. 2026Video-OPD: Efficient Post-Training of Multimodal Large Language Models for Temporal Video Grounding via On-Policy Distillationcited by 32
  26. 2025SenseNova-MARS: Empowering Multimodal Agentic Reasoning and Search via Reinforcement Learningcited by 31
  27. 2025JarvisEvo: Towards a Self-Evolving Photo Editing Agent with Synergistic Editor-Evaluator Optimizationcited by 30
  28. 2026Rethinking Video Generation Model for the Embodied Worldcited by 30
  29. 2026Vision-DeepResearch Benchmark: Rethinking Visual and Textual Search for Multimodal Large Language Modelscited by 30
  30. 2026LongCat-Next: Lexicalizing Modalities as Discrete Tokenscited by 30
  31. 2026Reward Hacking in the Era of Large Models: Mechanisms, Emergent Misalignment, Challengescited by 30
  32. 2026HiDream-O1-Image: A Natively Unified Image Generative Foundation Model with Pixel-level Unified Transformercited by 29
  33. 2026Qwen-Image-2.0 Technical Reportcited by 28
  34. 2026CLAP: Contrastive Latent Action Pretraining for Learning Vision-Language-Action Models from Human Videoscited by 27
  35. 2026MMFineReason: Closing the Multimodal Reasoning Gap via Open Data-Centric Methodscited by 27
  36. 2026StarVLA-α: Reducing Complexity in Vision-Language-Action Systemscited by 27
  37. 2026SenseNova-U1: Unifying Multimodal Understanding and Generation with NEO-unify Architecturecited by 27
  38. 2026VINO: A Unified Visual Generator with Interleaved OmniModal Contextcited by 26

50 most cited of 2,534 All 2,534 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2023Qwen-VL: A Versatile Vision-Language Model for Understanding, Localization, Text Reading, and Beyond
  2. 2024Enhancing Advanced Visual Reasoning Ability of Large Language Models
  3. 2026Firebolt-VL: Efficient Vision-Language Understanding with Cross-Modality Modulation
  4. 2025Qianfan-VL: Domain-Enhanced Universal Vision-Language Models
  5. 2025AndesVL Technical Report: An Efficient Mobile-side Multimodal Large Language Model
  6. 202620/20 Vision Language Models: A Prescription for Better VLMs through Data Curation Alone
  7. 2024Expanding Performance Boundaries of Open-Source Multimodal Models with Model, Data, and Test-Time Scaling
  8. 2024Beyond Visual Understanding: Introducing PARROT-360V for Vision Language Model Benchmarking