Meituan First-Quarter Losses Narrow; Wang Xing, Tencent Join Forces on AI Entry Point

Meituan has released its first-quarter 2026 earnings report, posting total revenue of 91 billion yuan, up 5.6% year over year. The operating loss narrowed significantly to 6.5 billion yuan from 16.1 billion yuan in the previous quarter—a reduction of nearly 10 billion yuan. Within that, losses from the core local commerce business plummeted from 10 billion yuan to 2 billion yuan, signaling a strong recovery in profitability. CEO Wang Xing noted that the food delivery sector is moving past intense competition toward a focus on efficiency and user experience. He expects the unit economics of food delivery to improve further in the second quarter, driven by seasonal demand.
Alongside the earnings release, Wang Xing announced a major AI partnership: Meituan's AI assistant "Xiaomei" will deeply integrate with Tencent's AI application "Yuanbao" and launch soon. After the collaboration, users can directly order food delivery and other local services through "Xiaomei" within the "Yuanbao" app, enabling a seamless closed-loop transaction. Wang Xing emphasized that serving AI agents (To A) will become as important as serving consumers and merchants. Meituan will continue to enhance its large model capabilities using its logistics system, merchant data, and user reviews. The company's self-developed trillion-parameter model, LongCat-2.0-Preview, has entered open testing, fully powered by domestic computing clusters.
Beyond software AI initiatives, Meituan has also advanced in hardware and low-altitude logistics. As of May this year, Meituan's drone-based "urban low-altitude network" has started regular deliveries in cities including Beijing, Shanghai, and Shenzhen, with cumulative commercial orders exceeding 900,000.
Interestingly, while Meituan continues to invest heavily in its own AI business, its equity investments in AI are already yielding returns. For the first time, Meituan publicly disclosed part of its investment portfolio: as of the end of the first quarter of 2026, the company holds 12.73% of Li Auto. In embodied intelligence and large models, Meituan owns 3.86% of Zhipu and 7.61% of Yuzhu Technology, which successfully passed its Sci-Tech Innovation Board listing review on the day the report was released. Additionally, Meituan fully acquired Guangnianzhiwai and led the early investment in Moonshot (Kimi). With capital markets increasingly bullish on embodied intelligence and large models, Meituan's early tech ecosystem bets are turning into substantial financial and strategic gains.
Related article
Why Doctors Struggle to Return Your Calls
While public discourse on AI in healthcare centers on diagnostics, drug discovery, and clinical consultations, a critical, often overlooked bottleneck determines whether patients ever receive care at all. This challenge stems less from physician shor
Anthropic’s Sonnet 5.5 Leak: 2M Token Context Challenges DeepSeek to Reclaim Cost-Effectiveness
Anthropic has unveiled its latest Sonnet 5.5 model, internally known as "Fennec" (after the desert fox), with a release expected next month. Leaked details indicate the model supports a 2 million token context window, offering faster reasoning, lower
Google DeepMind launches institute to widen the AGI debate
Google and Google DeepMind established the DeepMind Institute on Wednesday to foster dialogue on artificial general intelligence. The institute is led by DeepMind co-founder Shane Legg, Google executive James Manyika, and Google DeepMind chair Demis
Related Special Topic Recommendations
Comments (0)
0/500

Meituan has released its first-quarter 2026 earnings report, posting total revenue of 91 billion yuan, up 5.6% year over year. The operating loss narrowed significantly to 6.5 billion yuan from 16.1 billion yuan in the previous quarter—a reduction of nearly 10 billion yuan. Within that, losses from the core local commerce business plummeted from 10 billion yuan to 2 billion yuan, signaling a strong recovery in profitability. CEO Wang Xing noted that the food delivery sector is moving past intense competition toward a focus on efficiency and user experience. He expects the unit economics of food delivery to improve further in the second quarter, driven by seasonal demand.
Alongside the earnings release, Wang Xing announced a major AI partnership: Meituan's AI assistant "Xiaomei" will deeply integrate with Tencent's AI application "Yuanbao" and launch soon. After the collaboration, users can directly order food delivery and other local services through "Xiaomei" within the "Yuanbao" app, enabling a seamless closed-loop transaction. Wang Xing emphasized that serving AI agents (To A) will become as important as serving consumers and merchants. Meituan will continue to enhance its large model capabilities using its logistics system, merchant data, and user reviews. The company's self-developed trillion-parameter model, LongCat-2.0-Preview, has entered open testing, fully powered by domestic computing clusters.
Beyond software AI initiatives, Meituan has also advanced in hardware and low-altitude logistics. As of May this year, Meituan's drone-based "urban low-altitude network" has started regular deliveries in cities including Beijing, Shanghai, and Shenzhen, with cumulative commercial orders exceeding 900,000.
Interestingly, while Meituan continues to invest heavily in its own AI business, its equity investments in AI are already yielding returns. For the first time, Meituan publicly disclosed part of its investment portfolio: as of the end of the first quarter of 2026, the company holds 12.73% of Li Auto. In embodied intelligence and large models, Meituan owns 3.86% of Zhipu and 7.61% of Yuzhu Technology, which successfully passed its Sci-Tech Innovation Board listing review on the day the report was released. Additionally, Meituan fully acquired Guangnianzhiwai and led the early investment in Moonshot (Kimi). With capital markets increasingly bullish on embodied intelligence and large models, Meituan's early tech ecosystem bets are turning into substantial financial and strategic gains.
Why Doctors Struggle to Return Your Calls
While public discourse on AI in healthcare centers on diagnostics, drug discovery, and clinical consultations, a critical, often overlooked bottleneck determines whether patients ever receive care at all. This challenge stems less from physician shor
Anthropic’s Sonnet 5.5 Leak: 2M Token Context Challenges DeepSeek to Reclaim Cost-Effectiveness
Anthropic has unveiled its latest Sonnet 5.5 model, internally known as "Fennec" (after the desert fox), with a release expected next month. Leaked details indicate the model supports a 2 million token context window, offering faster reasoning, lower
Google DeepMind launches institute to widen the AGI debate
Google and Google DeepMind established the DeepMind Institute on Wednesday to foster dialogue on artificial general intelligence. The institute is led by DeepMind co-founder Shane Legg, Google executive James Manyika, and Google DeepMind chair Demis





Home






