According to a report by Sina Finance today, Meituan’s CEO said at the company’s second-quarter earnings call that AI represents a comprehensive upgrade of Meituan’s organization, products, and workflows.
Wang Xing said: “We have always believed that the long-term value of AI lies in whether it can truly enter real-world workflows, solve real-world problems, and develop into long-term, stable infrastructure capabilities. At the same time, fully domestic training and inference capabilities have helped us establish stronger advantages in long-term costs and infrastructure controllability.”
He continued: “We will not become a Token factory. Our models and AI products will support our core business, improving user and merchant experiences as well as the company’s operational efficiency.”
At the end of June this year, Meituan released its next-generation trillion-parameter large model, LongCat-2.0. The model is described as the industry’s first trillion-parameter model to complete end-to-end training and inference on a domestic computing cluster with 50,000 cards (1.6 T total parameters, approximately 48 B activated on average, with a dynamic range of 33B~56B). It was pretrained from scratch and natively supports an ultra-long context of 1M.