February’s delivery numbers for China’s new-energy carmakers underline a familiar pattern: growth is slowing, inventories are building and the market is shifting from headline volumes to product timing and margins. Leapmotor (零跑) retained the top spot among the new challengers with 28,067 deliveries, narrowly ahead of Li Auto (理想) at 26,421. NIO’s three-brand group delivered 20,797 vehicles, led by the full-size ES8, while Xiaomi, AITO (问界) and Xpeng lagged behind with Xiaomi and AITO each around the 18,000–20,000 mark and Xpeng at 15,256.
Seasonality and product transitions explain much of February’s weakness. Lunar New Year typically depresses monthly sales, and a wave of model changeovers has left several firms in a “product gap” where outgoing models are discounted and new ones are not yet in volume production. Financial incentives have moved from differentiation to table stakes: the so-called “seven-year low-interest” financing has become an industry norm and some manufacturers are even offering purchase-tax guarantees to sustain orders.
The numbers also reflect diverging strategic positions. Leapmotor’s relative stability—annualised growth and only a modest month-on-month fall—flows from deep penetration in the 100,000–200,000 yuan price band and a heavy emphasis on cost control through vertical integration. Li Auto’s resilience owes to steady demand for its extended-range models and aggressive end-of-life promotions on L-series vehicles as it prepares a mid-cycle refresh and launches the flagship L9 to burnish its “AI tech” credentials.
Other entrants paint a more worrying picture. Xiaomi’s deliveries more than halved month on month as the first-generation SU7 was retired and the company shifted factory capacity to a new SU7 due in April. AITO, marketed within Huawei’s HarmonyOS ecosystem and operated by Seres (赛力斯), saw February deliveries slump to roughly 18,000 after a January surge; the firm has doubled down on its higher-margin Wenjie line-up and is betting on a new M6 model to fill the crucial 230,000–280,000 yuan segment.
Xpeng’s shortfall highlights a broader industry trap: late-to-market products and narrow model funnels. The company has pushed an MPV and introduced range-extender variants for core sedans, while reorganising R&D around a single “general intelligence” centre to federate autonomous driving, cabin software and robotics. Yet those moves are capital intensive and their commercial payoff hinges on convincing 20-million-yuan-segment consumers that higher software and AD ambitions translate into practical value.
Legacy automakers and their EV sub-brands remain important. Geely sold 206,200 units in February, with Zeekr contributing 23,867; BYD moved 190,200 cars, including niche brands such as Fangchengbao (方程豹) and Denza; Great Wall delivered 72,594 with WEY posting 5,615. These incumbents provide ballast to the market and suggest that China’s growth is not collapsing, merely decelerating: the China Association of Automobile Manufacturers now forecasts 15.2% new-energy growth for 2026, down from the heady expansion of recent years.
What matters beyond the monthly statistics is what happens in March and the second quarter. A raft of new models is poised to hit showrooms—Leapmotor’s A10 and D19, Li Auto’s new L9, NIO’s ES9 and Ledo L80, Xiaomi’s upgraded SU7 and Xpeng’s range-extender versions. Consumer behaviour in China has hardened into “buy new, not old,” meaning the market reception of these launches will determine who can convert short-term incentives into sustainable share. Margins and cash flow will matter more than headline deliveries: raw-material inflation and rising component costs leave little room for prolonged discounting.
The industry is entering a more Darwinian phase. Firms with deep product pipelines, disciplined cost structures, credible supply partnerships and software that customers value will win in a slower-growth environment. Those dependent on finance-led demand or single-model volume will find the coming months difficult, and March’s calendar of introductions looks set to separate the contenders from the also-rans.