PAI-Bench: A Comprehensive Benchmark For Physical AI

TL;DR

PAI-Bench is a comprehensive benchmark designed to evaluate the perception and prediction capabilities of models in Physical AI, revealing current limitations in physical coherence and causal reasoning across models.

Contribution

Introduces PAI-Bench, a unified benchmark for assessing physical perception and prediction in models, covering video generation and understanding with real-world cases and specialized metrics.

Findings

Video generative models lack physical coherence despite visual fidelity.

Multi-modal large language models show limited forecasting and causal reasoning.

Current models are in early stages of handling Physical AI perceptual and predictive tasks.

Abstract

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.