FlagEval is a professional large-scale model evaluation platform created by the Beijing Academy of Artificial Intelligence, providing standardized quantitative assessment of the capabilities of AI models.