At Meituan's second-quarter earnings call on August 28, CEO Wang Xing said AI is driving a full upgrade of the company's organization, products and workflows, according to Jiemian News. He said Meituan sees the long-term value of AI in entering real work processes, solving real problems and building stable infrastructure capabilities, and added that domestically produced training and inference capabilities give the company an advantage in long-term costs and infrastructure controllability.
Wang said Meituan will not become a "token factory" and that its models and AI products will be used to support its core business, improve user and merchant experience, and raise operating efficiency. He said the company's AI strategy centers on three areas: building large language models, AI applications for work, and AI features in products.
On the model side, Meituan launched its next-generation self-developed model LongCat 2.0 in June. The company said it is the industry's first trillion-parameter model to complete both training and inference on a fully domestically produced computing cluster. Meituan said LongCat 2.0 has shown strong performance in code writing and agent-call scenarios and has been widely validated by the global developer community. On the product side, Meituan is upgrading its "Xiaotuan" AI assistant so AI can move from information search toward task execution, while strengthening end-to-end service capabilities in local-life scenarios. On the organizational side, Meituan said AI is being pushed deeper into workflows, and employees are now widely using AI tools in daily work.