Meituan's new AI Agent platform, already deployed across 90,000 employees with 30,000 agents in production, signals the food delivery giant's push to monetize enterprise AI beyond its core local services business.
Meituan (03690.HK) officially launched CatPaw, a full-scenario AI Agent platform, on July 27, offering enterprises a ready-to-use AI workspace and tools to build custom agents for daily operations. The platform has already been validated across catering, beauty services and pet hospitals — three sectors where Meituan holds deep operational data from its core food delivery and local services marketplace.
"CatPaw allows enterprises to build collaborative AI assistants that can handle complex workflows across multiple business scenarios, from customer service to inventory management," a Meituan representative said. The platform provides enterprise-grade agent development and hosting capabilities, enabling merchants to create AI assistants without specialized machine learning expertise.
The launch builds on Meituan's broader AI infrastructure push. In June, the company unveiled LongCat-2.0, a 1.6-trillion-parameter large language model built on a Mixture of Experts architecture that activates roughly 48 billion parameters per token. The model natively supports a 1-million-token context window — enough to process inputs of up to a million words at once — and is the industry's first trillion-parameter model to complete training and inference entirely on domestic Chinese computing power. Meituan also denied market rumors that Pei Peng, head of foundational models for the LongCat team, was planning to leave the company.
Why CatPaw matters for Meituan's AI strategy
Meituan's internal deployment numbers — 90,000 employees using the platform, 30,000 agents created — suggest the company is treating AI as an operational efficiency lever before selling it externally. The platform's validation across real business scenarios gives Meituan a data advantage over pure-play AI vendors that lack domain-specific training data in local services.
The competitive stakes are high. Alibaba's Tongyi Qianwen and Tencent's Hunyuan models are also targeting enterprise AI workloads in China, and both companies have deeper cloud infrastructure businesses to cross-sell AI services. Meituan's edge lies in its vertical-specific data: the company processes millions of daily transactions across food delivery, restaurant bookings, beauty appointments and pet care, giving it training data that general-purpose AI models cannot easily replicate.
Stock market reaction and investor implications
Meituan shares rose 4.21% to HK$90.35 on launch day, with turnover reaching HK$3.76 billion (about $480 million). Short selling accounted for HK$807 million, or 21.14% of total volume, indicating significant hedging activity around the announcement. The stock has gained roughly 12% year to date, according to exchange data.
Analysts at Guotai Haitong Securities identified AI-powered automated fulfillment as one of three growth drivers — alongside instant retail and the overseas KeeTa business — that could provide valuation re-rating opportunities for Meituan. The company is deploying autonomous delivery vehicles and embodied robots alongside its large language models, positioning itself as a service-industry technology operator rather than a pure transaction platform.
For investors, the key question is whether CatPaw can generate meaningful revenue beyond internal cost savings. Chinese AI model vendors are under pressure to show returns on massive infrastructure spending: Z.ai reported revenue of 724 million yuan ($107 million) last year but a net loss of 4.7 billion yuan ($694 million). Meituan's advantage is that its AI investments can be justified by internal efficiency gains alone, reducing the pressure to monetize externally. The company trades at roughly 18x forward earnings, according to Bloomberg consensus estimates, with the AI segment not yet a material contributor to revenue.
This article is for informational purposes only and does not constitute investment advice.