Meituan (美团) posts 2025 results — losses narrow, cash pile holds, but AI push raises new questions

Solid headline numbers, but is the hard part over?

Meituan (美团) reported 2025 full-year results showing resilience after a bruising food-delivery price war. Revenue reached 3,649亿元, up 8.1% year‑on‑year, while net loss came in at 234亿元 — inside the company’s earlier guidance range and better than some feared. Core local commerce losses narrowed sharply in Q4 to about 100亿元 from 141亿元 in Q3. Meituan ended the year with sizeable liquidity: cash of 1,068亿元 and short‑term investments of 601亿元, signaling it still has "ammunition" to defend market share.

That performance comes against a backdrop of intense competition. It has been reported that Alibaba (阿里巴巴), JD (京东) and Meituan — along with new entrants such as Taobao Flash Sale (淘宝闪购) and short‑video platforms — poured roughly 800–1,000亿元 into subsidies during the battle for users. Beijing’s market regulator even re‑published an Economic Daily commentary urging an end to the subsidy war, saying healthy rivalry should come through technology and efficiency rather than price dumping. But does a regulatory nudge mean Meituan can relax? Not necessarily — rivals are probing its core local services, and the competitive window remains open.

A full‑throated AI strategy — expensive and urgent

CEO Wang Xing has framed Meituan’s response as offensive: “In the AI revolution, the only reasonable strategy is attack, not defense,” he told investors. Meituan has accelerated spending — R&D rose to 260亿元 (up 23.5% and lifting R&D intensity to 7.1% of revenue), with Q4 alone at about 70亿元 — and publicly promoted its in‑house LongCat large model, a three‑layer roadmap for “AI‑powered apps,” and consumer‑facing assistants such as “Xiaomei” and the in‑app “Xiaotuan.” It has been reported that Meituan claims its AI outlay may be among the largest of any Chinese internet firm outside of cloud incumbents.

Wang and Meituan executives emphasize a differentiated route: embed AI into the company’s physical‑world logistics and service network — millions of couriers, hundreds of thousands of merchants, and assets like drones and delivery robots — rather than competing head‑on in generic large models. The pitch to investors is simple: apply AI to lift efficiency across a heavy‑asset, local‑service empire.

Big opportunity, big risks

That thesis has logic, but execution gaps are visible. Meituan’s AI gains on the B‑side — tools for merchants that boost efficiency and conversion — show promising metrics (it has been reported that some AI tools improved site‑selection success and customer‑service efficiency markedly). But consumer (C‑end) AI products remain fragmented, with multiple assistants and experiments failing to cohere into a clear mainstream offering. Some internal projects, such as the Tabbit AI browser, reportedly drew controversy and have not gelled with core business flows.

There is a strategic trade‑off: Meituan is trimming non‑core expansion, refocusing on profitability and calibrating headcount for overseas plans, while doubling down on AI. But if third‑party AI agents from the likes of Alibaba, ByteDance (字节跳动), Baidu (百度) or others become the primary consumer decision layer, Meituan risks becoming a commoditized “service pipeline” unless its AI can deliver true user‑facing stickiness. For Western readers: this is not just a corporate technology bet but a test of platform power in China’s fast‑shifting regulatory and competitive landscape. Meituan’s numbers show recovery; its next challenge is turning heavy AI spending into defensible, consumer‑facing advantage.