# PAI-Bench: A Comprehensive Benchmark For Physical AI

arXiv: https://arxiv.org/abs/2512.01989 · HTML: https://mlsanity.com/paper/2512.01989/

## Key facts

- Cited by 34 papers in the map (data through 14 Oct 2026).
- Placed in Embodied and 3D vision-language models, one of 16 directions in Multimodal LLMs / vision-language models.
- Builds on 83 of its 101 references that are in the map.
- First posted on arXiv 1 Dec 2025.

## Placed in

- Multimodal LLMs / vision-language models / [Embodied and 3D vision-language models](https://mlsanity.com/mllm/embodied-and-3d-vision-language-models/)

## Builds on (most cited first, 30 of 83)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [The Unreasonable Effectiveness of Deep Features as a Perceptual Metric](https://mlsanity.com/paper/1801.03924/) (2018)
- [Qwen3 Technical Report](https://mlsanity.com/paper/2505.09388/) (2025)
- [Visual Instruction Tuning](https://mlsanity.com/paper/2304.08485/) (2023)
- [Emerging Properties in Self-Supervised Vision Transformers](https://mlsanity.com/paper/2104.14294/) (2021)
- [Efficient Memory Management for Large Language Model Serving with PagedAttention](https://mlsanity.com/paper/2309.06180/) (2023)
- [Qwen2.5-VL Technical Report](https://mlsanity.com/paper/2502.13923/) (2025)
- [GPT-4o System Card](https://mlsanity.com/paper/2410.21276/) (2024)
- [Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution](https://mlsanity.com/paper/2409.12191/) (2024)
- [Grounding DINO: Marrying DINO with Grounded Pre-Training for Open-Set Object Detection](https://mlsanity.com/paper/2303.05499/) (2023)
- [DreamBooth: Fine Tuning Text-to-Image Diffusion Models for Subject-Driven Generation](https://mlsanity.com/paper/2208.12242/) (2022)
- [SAM 2: Segment Anything in Images and Videos](https://mlsanity.com/paper/2408.00714/) (2024)
- [Wan: Open and Advanced Large-Scale Video Generative Models](https://mlsanity.com/paper/2503.20314/) (2025)
- [Making the V in VQA Matter: Elevating the Role of Image Understanding in Visual Question Answering](https://mlsanity.com/paper/1612.00837/) (2016)
- [CogVideoX: Text-to-Video Diffusion Models with An Expert Transformer](https://mlsanity.com/paper/2408.06072/) (2024)
- [Qwen-VL: A Versatile Vision-Language Model for Understanding, Localization, Text Reading, and Beyond](https://mlsanity.com/paper/2308.12966/) (2023)
- [VBench: Comprehensive Benchmark Suite for Video Generative Models](https://mlsanity.com/paper/2311.17982/) (2023)
- [HunyuanVideo: A Systematic Framework For Large Video Generative Models](https://mlsanity.com/paper/2412.03603/) (2024)
- [World Models](https://mlsanity.com/paper/1803.10122/) (2018)
- [InternVL3.5: Advancing Open-Source Multimodal Models in Versatility, Reasoning, and Efficiency](https://mlsanity.com/paper/2508.18265/) (2025)
- [Depth Anything V2](https://mlsanity.com/paper/2406.09414/) (2024)
- [Video-MME: The First-Ever Comprehensive Evaluation Benchmark of Multi-modal LLMs in Video Analysis](https://mlsanity.com/paper/2405.21075/) (2024)
- [Depth Map Prediction from a Single Image using a Multi-Scale Deep Network](https://mlsanity.com/paper/1406.2283/) (2014)
- [MVBench: A Comprehensive Multi-modal Video Understanding Benchmark](https://mlsanity.com/paper/2311.17005/) (2023)
- [MUSIQ: Multi-scale Image Quality Transformer](https://mlsanity.com/paper/2108.05997/) (2021)
- [How Far Are We to GPT-4V? Closing the Gap to Commercial Multimodal Models with Open-Source Suites](https://mlsanity.com/paper/2404.16821/) (2024)
- [CogVideo: Large-scale Pretraining for Text-to-Video Generation via Transformers](https://mlsanity.com/paper/2205.15868/) (2022)
- [HowTo100M: Learning a Text-Video Embedding by Watching Hundred Million Narrated Video Clips](https://mlsanity.com/paper/1906.03327/) (2019)
- [WorldSimBench: Towards Video Generation Models as World Simulators](https://mlsanity.com/paper/2410.18072/) (2024)
- [Cosmos World Foundation Model Platform for Physical AI](https://mlsanity.com/paper/2501.03575/) (2025)

## Built on by (most cited first, 30 of 34)

- [Cosmos 3: Omnimodal World Models for Physical AI](https://mlsanity.com/paper/2606.02800/) (2026)
- [ABot-PhysWorld: Interactive World Foundation Model for Robotic Manipulation with Physics Alignment](https://mlsanity.com/paper/2603.23376/) (2026)
- [RoboWM-Bench: A Benchmark for Evaluating World Models in Robotic Manipulation](https://mlsanity.com/paper/2604.19092/) (2026)
- [SpatialClaw: Rethinking Action Interface for Agentic Spatial Reasoning](https://mlsanity.com/paper/2606.13673/) (2026)
- [Self-Refining Video Sampling](https://mlsanity.com/paper/2601.18577/) (2026)
- [Code as Worlds: Agentic Discovery of Executable World Representations for Physical Reasoning](https://mlsanity.com/paper/2608.27549/) (2026)
- [Out of Sight, Out of Mind? Evaluating State Evolution in Video World Models](https://mlsanity.com/paper/2603.13215/) (2026)
- [PhysisForcing: Physics Reinforced World Simulator for Robotic Manipulation](https://mlsanity.com/paper/2606.28128/) (2026)
- [Kairos: A Regret-Aware Native World-Action Model Stack for Physical AI](https://mlsanity.com/paper/2606.16533/) (2026)
- [CausalSpatial: A Benchmark for Object-Centric Causal Spatial Reasoning](https://mlsanity.com/paper/2601.13304/) (2026)
- [RoboTrustBench: Benchmarking the Trustworthiness of Video World Models for Robotic Manipulation](https://mlsanity.com/paper/2606.01600/) (2026)
- [OmniFysics: Towards Physical Intelligence Evolution via Omni-Modal Signal Processing and Network Optimization](https://mlsanity.com/paper/2602.07064/) (2026)
- [Vesta: A Generalist Embodied Reasoning Model](https://mlsanity.com/paper/2606.20905/) (2026)
- [WorldCache: Content-Aware Caching for Accelerated Video World Models](https://mlsanity.com/paper/2603.22286/) (2026)
- [Reward as An Agent for Embodied World Models](https://mlsanity.com/paper/2606.19990/) (2026)
- [Benchmarking Single-Factor Physical Video-to-Audio Generation](https://mlsanity.com/paper/2605.30339/) (2026)
- [RxBrain: Embodied Cognition Foundation Model with Joint Language-Visual Reasoning and Imagination](https://mlsanity.com/paper/2607.14187/) (2026)
- [RoboGaze: Evaluating Robot World Models via Structured Vision-Language Analysis](https://mlsanity.com/paper/2606.28385/) (2026)
- [CausalWM: Causal Chain-of-Thought Reasoning for Embodied World Model](https://mlsanity.com/paper/2609.23184/) (2026)
- [PhysVista: Benchmarking Physical Intelligence in VLMs via a Perception-Reasoning-Assessment Loop](https://mlsanity.com/paper/2610.00559/) (2026)
- [ChronoPhyBench: Do MLLMs Truly Understand the World or Merely Exploit Language Priors?](https://mlsanity.com/paper/2606.07962/) (2026)
- [REALM: A Unified Red-Teaming Benchmark for Physical-World VLMs](https://mlsanity.com/paper/2606.23892/) (2026)
- [TrustRoboReward: Preference-Ordered Isotonic Score Editing for Multi-Paradigm Robot Reward Models](https://mlsanity.com/paper/2608.08491/) (2026)
- [VLMs Win a Systematic Evaluation of Underwater Image Reconstruction](https://mlsanity.com/paper/2608.11425/) (2026)
- [PhysAI-Bench: A Benchmark for LLM-Based Agentic Decision-Making in Autonomous UAV-Centric Physical AI](https://mlsanity.com/paper/2609.23695/) (2026)
- [OmniFysics-Nano-V2 Technical Report: Understanding the Physical World Across Modalities](https://mlsanity.com/paper/2609.25738/) (2026)
- [DyMD: Preserving Interaction Dynamics through Distribution Matching Distillation in Few-Step Video World Models](https://mlsanity.com/paper/2609.31349/) (2026)
- [LongLive-Plug: Once-for-All Distillation for Video Generation](https://mlsanity.com/paper/2609.38154/) (2026)
- [PhyProbe: Rethinking Physical Consistency Evaluation in Generated Videos](https://mlsanity.com/paper/2609.38377/) (2026)
- [WOVEN: Weaving Visual World Modeling into Multimodal LLMs](https://mlsanity.com/paper/2610.12417/) (2026)

## Abstract (arXiv)

Physical AI aims to develop models that can perceive and predict real-world dynamics; yet, the extent to which current multi-modal large language models and video generative models support these abilities is insufficiently understood. We introduce Physical AI Bench (PAI-Bench), a unified and comprehensive benchmark that evaluates perception and prediction capabilities across video generation, conditional video generation, and video understanding, comprising 2,808 real-world cases with task-aligned metrics designed to capture physical plausibility and domain-specific reasoning. Our study provides a systematic assessment of recent models and shows that video generative models, despite strong visual fidelity, often struggle to maintain physically coherent dynamics, while multi-modal large language models exhibit limited performance in forecasting and causal interpretation. These observations suggest that current systems are still at an early stage in handling the perceptual and predictive demands of Physical AI. In summary, PAI-Bench establishes a realistic foundation for evaluating Physical AI and highlights key gaps that future systems must address.