Xiaomi’s profit drops 43%
Xiaomi Corp.’s adjusted net profit fell 43.1% in the first quarter, missing analysts’ expectations as surging memory-chip costs, weaker smartphone shipments and slower electric-vehicle deliveries weighed on the Hong Kong-listed company. Revenue fell 10.9% from a year earlier to 99.14 billion yuan ($14.6 billion), below the LSEG consensus estimate of 103.4 billion yuan. Adjusted net profit came in at 6.07 billion yuan, short of the 6.4 billion yuan forecast. Smartphone revenue, which accounts for nearly half of total sales, fell 12.5% to 44.3 billion yuan. President Lu Weibing said Xiaomi had deliberately reduced inventory of low- and mid-end devices in distribution channels. Auto revenue totaled 19 billion yuan in the quarter. The average selling price fell to 235,000 yuan from 250,000 yuan in the fourth quarter, reflecting purchase-tax subsidies and lower-priced inventory clearance.
Alipay’s AI payment transactions top 300 million
Ant Group’s Alipay announced on Tuesday that its “AI Pay” system has processed 300 million agent-driven transactions, claiming it as the world’s first large-scale commercial AI-native payment infrastructure. The platform also introduced an AI wallet that allows users to manage agent-based tasks in real time before and during payments, alongside post-payment billing queries. Additionally, Alipay launched Token Pay, a comprehensive solution to help large language model developers manage global user subscriptions and token top-ups.
Pony.ai’s robotaxi revenue hits record
Autonomous driving startup Pony.ai reported first-quarter total revenue of 236 million yuan, a 145% year-on-year increase. Its robotaxi business generated 59 million yuan, surging 395.4% from a year earlier. Operating a fleet of over 1,700 robotaxis as of May, the company raised its full-year targets for the segment, projecting 2026 revenue to exceed 3.5 times the 2025 level and expanding its fleet to over 3,500 vehicles across more than 20 cities globally.
Baichuan unveils new medical AI model
AI startup Baichuan introduced its new-generation medical large language model, Baichuan-M4, alongside an AI family doctor application. Founder and CEO Wang Xiaochuan stated that practical AI applications in healthcare require specialized models with low hallucination rates, strong evidence-based reasoning and proactive questioning capabilities, which general models currently lack. Wang claimed the new model has reduced its factual hallucination rate to a globally unprecedented low of 3.3%.
