MORE: multilingual document parsing benchmark covering 149 languages

Researchers introduce MORE, a large-scale benchmark designed to evaluate the performance of Vision-Language Models on multilingual document parsing. The dataset addresses the lack of ground truth for low-resource languages by curating samples from real-world documents through a model-assisted, human-refined annotation pipeline.

• Covers 149 languages, making it the most linguistically diverse benchmark to date.
• Extends evaluation beyond plain text to include structural elements such as code blocks, tables, and catalogs.
• Establishes new performance baselines for long-tail languages using state-of-the-art models.

The benchmark aims to diagnose model capabilities in realistic, diverse scenarios where existing tools predominantly focus on high-resource languages like English and Chinese.