OpenCompass is a professional large-scale model evaluation system developed by the Shanghai Artificial Intelligence Laboratory. It aims to quantify the actual capabilities of AI models through a standardized evaluation process.