Awesome Benchmarks

Key papers

• DreamX-World 1.0: A General-Purpose Interactive World Model (2026)
  DreamX Team et al.
  15.33
• InterleaveThinker: Reinforcing Agentic Interleaved Generation (2026)
  Dian Zheng et al.
  14.38
• VibeThinker-3B: Exploring the Frontier of Verifiable Reasoning in Small Language Models (2026)
  Sen Xu et al.
  14.06
• LabVLA: Grounding Vision-Language-Action Models in Scientific Laboratories (2026)
  Baochang Ren et al.
  13.59
• STEP3-VL-10B Technical Report (2026)
  Ailin Huang et al.
  13.07
• Urban Socio-Semantic Segmentation with Vision-Language Reasoning (2026)
  Yu Wang et al.
  12.58
• Orchestra-o1: Omnimodal Agent Orchestration (2026)
  Fan Zhang et al.
  12.21
• VidVec: Unlocking Video MLLM Embeddings for Video-Text Retrieval (2026)
  Issar Tzachor et al.
  12.16
• Vision-DeepResearch Benchmark: Rethinking Visual and Textual Search for Multimodal Large Language Models (2026)
  Yu Zeng et al.
  12.05
• Geometric Action Model for Robot Policy Learning (2026)
  Jisang Han et al.
  11.96
• Everything in Its Place: Benchmarking Spatial Intelligence of Text-to-Image Models (2026)
  Zengbin Wang et al.
  11.87
• FORGE: Fine-grained Multimodal Evaluation for Manufacturing Scenarios (2026)
  Xiangru Jian et al.
  11.70
• M$^3$Eval: Multi-Modal Memory Evaluation through Cognitively-Grounded Video Tasks (2026)
  Jie Huang et al.
  11.47
• Task-Focused Memorization for Multimodal Agents (2026)
  Tao Zou et al.
  11.25
• VisualClaw: A Real-Time, Personalized Agent for the Physical World (2026)
  Haoqin Tu et al.
  11.22
• Think in Strokes, Not Pixels: Process-Driven Image Generation via Interleaved Reasoning (2026)
  Lei Zhang et al.
  11.11
• Qwen-RobotWorld Technical Report: Unifying Embodied World Modeling through Language-Conditioned Video Generation (2026)
  Jie Zhang et al.
  11.03
• SpecEyes: Accelerating Agentic Multimodal LLMs via Speculative Perception and Planning (2026)
  Haoyu Huang et al.
  10.74
• Zooming without Zooming: Region-to-Image Distillation for Fine-Grained Multimodal Perception (2026)
  Lai Wei et al.
  10.72
• TwinBrainVLA: Unleashing the Potential of Generalist VLMs for Embodied Tasks via Asymmetric Mixture-of-Transformers (2026)
  Bin Yu et al.
  10.59
• OmniVideo-100K: A Dataset for Audio-Visual Reasoning through Structured Scripts and Evidence Chains (2026)
  Xinyue Cai et al.
  10.59
• The Script is All You Need: An Agentic Framework for Long-Horizon Dialogue-to-Cinematic Video Generation (2026)
  Chenyu Mu et al.
  10.41
• MM-Zero: Self-Evolving Multi-Model Vision Language Models From Zero Data (2026)
  Zongxia Li et al.
  10.41
• Cross-Modal Retrieval for Motion and Text via DropTriple Loss (2023)
  Sheng Yan et al.
  10.35
• LLaVE: Large Language and Vision Embedding Models with Hardness-Weighted Contrastive Learning (2025)
  Zhibin Lan et al.
  10.31
• VideoDetective: Clue Hunting via both Extrinsic Query and Intrinsic Relevance for Long Video Understanding (2026)
  Ruoliu Yang et al.
  10.24
• UniCorn: Towards Self-Improving Unified Multimodal Models through Self-Generated Supervision (2026)
  Ruiyan Han et al.
  10.17
• Text-Vision Co-Instructed Image Editing (2026)
  Chenxi Xie et al.
  9.90
• CODA-BENCH: Can Code Agents Handle Data-Intensive Tasks? (2026)
  Yuxin Zhang et al.
  9.73
• VisGym: Diverse, Customizable, Scalable Environments for Multimodal Agents (2026)
  Zirui Wang et al.
  9.71
• SkinFlow: Efficient Information Transmission for Open Dermatological Diagnosis via Dynamic Visual Encoding and Staged RL (2026)
  Lijun Liu et al.
  9.65
• Watch Before You Answer: Learning from Visually Grounded Post-Training (2026)
  Yuxuan Zhang et al.
  9.65
• GutenOCR: A Grounded Vision-Language Front-End for Documents (2026)
  Hunter Heidenreich et al.
  9.54
• VideoAuto-R1: Video Auto Reasoning via Thinking Once, Answering Twice (2026)
  Shuming Liu et al.
  9.48
• Omnivec2 -- A Novel Transformer Based Network For Large Scale Multimodal And Multitask Learning (2025)
  Siddharth Srivastava, Gaurav Sharma
  9.46
• UniDDT: Unifying Multimodal Understanding and Generation with Decoupled Diffusion Transformer (2026)
  Shuai Wang et al.
  9.34
• Pushupbench: Your VLM Is Not Good At Counting Pushups (2026)
  Shengzhi Li, Jiarun Chen, Karun Sharma, et al.
  9.22
• RepWAM: World Action Modeling with Representation Visual-Action Tokenizers (2026)
  Junke Wang et al.
  9.19
• PhoneHarness: Harnessing Phone-Use Agents through Mixed GUI, CLI, and Tool Actions (2026)
  Chenxin Li et al.
  9.11
• VINO: A Unified Visual Generator with Interleaved OmniModal Context (2026)
  Junyi Chen et al.
  9.10
• X-SAM: From Segment Anything to Any Segmentation (2025)
  Hao Wang et al.
  9.02
• Multimodal Fake News Detection: MFND Dataset And Shallow-deep Multitask Learning (2025)
  Ye Zhu, Yunan Wang, Zitong Yu
  9.01
• DualToken: Towards Unifying Visual Understanding and Generation with Dual Visual Vocabularies (2025)
  Wei Song et al.
  9.00
• Go To Zero: Towards Zero-shot Motion Generation With Million-scale Data (2025)
  Ke Fan, Shunlin Lu, Minyue Dai, et al.
  8.97
• OS-Symphony: A Holistic Framework for Robust and Generalist Computer-Using Agent (2026)
  Bowen Yang et al.
  8.96
• A Safety Report on GPT-5.2, Gemini 3 Pro, Qwen3-VL, Grok 4.1 Fast, Nano Banana Pro, and Seedream 4.5 (2026)
  Xingjun Ma et al.
  8.81
• ClinHallu: A Benchmark for Diagnosing Stage-Wise Hallucinations in Medical MLLM Reasoning (2026)
  Sicheng Yang et al.
  8.47
• ViTextVQA: A Large-Scale Visual Question Answering Dataset and a Novel Multimodal Feature Fusion Method for Vietnamese Text Comprehension in Images (2024)
  Quan Van Nguyen et al.
  8.37
• Seeing Isn't Knowing: Do VLMs Know When Not to Answer Spatial Questions (and Why)? (2026)
  Yue Zhang et al.
  8.18
• RANKVIDEO: Reasoning Reranking for Text-to-Video Retrieval (2026)
  Tyler Skow et al.
  8.11
• Know "No" Better: A Data-Driven Approach for Enhancing Negation Awareness in CLIP (2025)
  Junsung Park et al.
  8.07
• MVEB: Massive Video Embedding Benchmark (2026)
  Adnan El Assadi et al.
  7.98
• DecAlign: Hierarchical Cross-Modal Alignment for Decoupled Multimodal Representation Learning (2025)
  Chengxuan Qian et al.
  7.82
• What Users Leave Unsaid: Under-Specified Queries Limit Vision-Language Models (2026)
  Dasol Choi et al.
  7.81
• Dream-VL & Dream-VLA: Open Vision-Language and Vision-Language-Action Models with Diffusion Language Model Backbone (2025)
  Jiacheng Ye et al.
  7.64
• Is CLIP ideal? No. Can we fix it? Yes! (2025)
  Raphi Kang et al.
  7.55
• Loc3R-VLM: Language-based Localization and 3D Reasoning with Vision-Language Models (2026)
  Kevin Qu et al.
  7.51
• VisPhyWorld: Probing Physical Reasoning via Code-Driven Video Reconstruction (2026)
  Jiarong Liang et al.
  7.45
• BEAR: Benchmarking and Enhancing Multimodal Language Models for Atomic Embodied Capabilities (2025)
  Yu Qi et al.
  7.41
• Semirnet: A Semantic Irony Recognition Network For Multimodal Sarcasm Detection (2025)
  Jingxuan Zhou, Yuehao Wu, Yibo Zhang, et al.
  7.41