Artifact-Bench: Evaluating MLLMs on Detecting and Assessing the Artifacts of AI-Generated Videos
For researchers and developers of video generative models, this benchmark reveals that current MLLMs are unreliable for evaluating AI-generated video realism.
Artifact-Bench evaluates 19 MLLMs on detecting and assessing artifacts in AI-generated videos across photorealistic, animated, and CG-style domains. Results show many models perform near random or below random, with significant misalignment to human judgments.
Recent video generative models have greatly improved the realism of AI-generated videos, yet their outputs still exhibit artifacts such as temporal inconsistencies, structural distortions, and semantic incoherence. While Multimodal Large Language Models (MLLMs) show strong visual understanding capabilities, their ability to perceive and reason about such artifacts remains unclear. Existing benchmarks often lack systematic evaluation of artifact-aware perception and fine-grained diagnostic reasoning, especially across diverse AI-generated video domains beyond photorealistic content. To address this gap, we introduce Artifact-Bench, a comprehensive benchmark for evaluating MLLMs on AI-generated video artifact detection and analysis. We first establish a three-level hierarchical taxonomy of realism artifacts, covering photorealistic, animated, and CG-style videos. Based on this taxonomy, Artifact-Bench defines three complementary tasks: real vs. AI-generated video classification, pairwise realism comparison, and fine-grained artifact identification. Experiments on 19 leading MLLMs reveal substantial limitations in artifact perception and reasoning, with many models approaching random or even below-random performance in challenging settings. We further observe significant misalignment between MLLM judgments and human perceptual preferences, highlighting their limited reliability as general evaluators for AI-generated video realism.