New Energy Vehicle Makers Enter Humanoid Robotics En Masse

August 2026 — dubbed the "Robotics Month" by industry insiders — saw Unitree Technology list on the capital market, Xiaomi unveil its first humanoid robot, and the second World Humanoid Robot Games draw over 600 teams from 16 countries across six continents. Yet the most telling signal came from China's leading new-energy vehicle (NEV) makers: they are collectively entering the embodied AI and humanoid robotics domain with unprecedented intensity.

Funding Breaks Records; Automakers Become Key 'Humanoid Builders'

On August 24, Xpeng Group announced the successful close of the first funding round for its humanoid robotics business, led by IDG Capital and co-invested by GSR Ventures, with Tencent and Alibaba joining as strategic investors. The round raised over $900 million, valuing the business at more than $6.3 billion post-money — the highest single-round private equity financing in China's embodied AI sector to date. Notably, although the group has rebranded, its core business remains focused on intelligent electric vehicles.

On the same day, NIO announced that Senior Vice President and Head of Intelligent Driving, Shaoqing Ren, has launched a startup focused on "Physical AI foundational models" and "embodied intelligence." NIO will serve as a strategic shareholder and engage in deep collaboration with the new venture. Leapmotor Automobile revealed during its 2026 mid-year earnings briefing that it has already planned entry into the embodied intelligent robotics field, with concrete details expected soon.

Li Automobile, meanwhile, declared its full pivot to embodied intelligent robotics R&D as early as 2025. Li Xiang reiterated during its Q2 2026 earnings call that its in-house battery and chip development strategy stems directly from their foundational technological value in embodied intelligence.

Why Automakers? Natural Capabilities Are Highly Transferable

Elon Musk, CEO of Tesla, forecasts that global humanoid robot units may reach 30–50 billion — three to five times Earth's human population. Meanwhile, Stanley Research estimates the global humanoid robot market could generate $7.5 trillion annually by 2050 — compared to the $4.36 trillion global automotive industry size in 2025.

Automakers hold inherent advantages rooted in capability reuse: AI technologies developed for autonomous driving — such as LiDAR, multimodal perception, and end-to-end large models — can be directly applied to robotic environmental sensing and decision-making. Core components — including electric drive systems, drive-by-wire chassis, battery management systems (BMS), servo motors, and gear reducers — also offer strong cross-platform adaptability. Furthermore, automakers' vehicle-level data closed-loop systems, testing & validation infrastructure, manufacturing scale, and real-world deployment experience all constitute critical enablers for robotics commercialization.

As Li Xiang put it: "Autonomous driving is the first half of embodied intelligence; general-purpose humanoid robots represent the second half."

Bubbles Emerge — Mass Production and Commercialization Remain the Biggest Bottlenecks

Beneath the hype, reality remains sobering. According to data from China's Hundred-Person Council on Humanoid Robots and Embodied Intelligence, only 41,200 units shipped globally in H1 2026 — roughly equivalent to the monthly sales volume of some top-tier NEV startups. Of these, approximately 65% were deployed for entertainment, demonstration, or education/research purposes; just 35% entered industrial production lines or logistics warehousing, while the consumer (C-end) market remains virtually untapped.

Yet investor enthusiasm continues to surge: global humanoid-robot-related funding reached ¥104.1 billion ($14.5B) in H1 2026 — exceeding the full-year 2025 total. In China alone, 105 deals raised ¥71 billion ($9.9B). Unitree Technology's market cap soared to ¥450 billion ($62.7B) on its first trading day — but shed over ¥200 billion ($27.9B) within days. Its 2025 revenue stood at just ¥1.7 billion ($237M) — less than 1/250th of its peak valuation.

Technically, high-end humanoid robots still face multiple hurdles: dynamic balance control, dexterous manipulation, force feedback, and fine-grained operation. Xpeng Group Chairman He Xiaopeng admitted that R&D difficulty for humanoid robots is "at least 20 times greater" than for intelligent vehicles. With standards undefined, technical roadmaps unsettled, and markets fragmented, viable commercial pathways remain highly uncertain.