90% & Rising: China Dominates the Humanoid Robot Race While the West Scrambles to Catch Up

01 — The Numbers Are Not Close: China's 90% Market Dominance in 2025

When the final tallies for 2025 global humanoid robot sales were compiled by Omdia and IDC, they revealed a market landscape that surprised even analysts who had been tracking China's robotics momentum closely. Between 13,000 and 18,000 humanoid robots were sold globally in 2025. Nearly 90% of them were Chinese.

Six of the highest-selling humanoid companies in the world were headquartered in China. The West, relying heavily on venture capital, high-profile demonstrations, and media-saturated announcements, was barely competing at the unit volume level.

This is not a story about Chinese robots being imitations of Western designs. It is a story about China executing a coherent national strategy. This framework combines state policy mandates, deep manufacturing infrastructure, mature component supply chains, aggressive AI investment, and government-driven demand from state-owned enterprises. They accomplished this faster and at a greater scale than any Western competitor expected.

The numbers illuminate a sharp distinction between the robotics market and the AI model market. In large language models, the U.S. maintained a clear lead through 2025. Yet in the physical embodiment of AI — where models interact with the real world through actuators, joints, and power systems — China's manufacturing advantages translated directly to market dominance. The robot is a machine. China knows how to build machines at scale.

02 — Why China Won the First Round: Policy, Supply Chain & State Demand

Lian Jye Su, chief analyst at Omdia, identified three interlocking forces driving China's early advantage. First is policy support and public investment. The Chinese government designated humanoid robotics a national strategic priority. Substantial direct subsidies, preferential access to manufacturing zones, and technology transfer programs channeled resources toward domestic champions. This playbook mirrors the strategies deployed in solar panels, electric vehicles, and batteries.

The second force is a mature and deep supply chain. The critical hardware components — electric motors, precision actuators, cameras, control electronics — are overwhelmingly manufactured in Asia. Chinese robotics companies enjoy proximity to their suppliers, resulting in shorter lead times, lower logistics costs, and the ability to co-develop components rapidly. Conversely, Western companies building humanoids frequently import key components from Asia at higher costs with greater supply chain vulnerability.

The third force is state-driven demand. Chinese state-owned enterprises across manufacturing, logistics, construction, and infrastructure have actively piloted and deployed humanoid robots. Government directives create a captive domestic market without a U.S. or European equivalent. Chinese manufacturers iterate faster because they have guaranteed buyers ready to deploy early-stage products.

03 — The Western Counterpunch: Software, AI, and the Optimus Gamble

What the West lacks in manufacturing scale, it aims to offset with superior AI architecture. The hardware of humanoid robots is currently converging toward commodity-level pricing across all major manufacturers. The differentiating variable over the next decade will be which robot houses the most capable AI — the ability to process natural language, operate autonomously in unstructured environments, and learn from experience.

Elon Musk framed Optimus not primarily as a hardware product but as an AI platform. Tesla intends to leverage billions of miles of real-world driving data collected by its vehicle fleet to build physical world models. Optimus Gen 3 is currently in its production-intent prototype phase with a 25-actuator-per-hand design. Tesla is targeting an internal-factory-first deployment strategy aimed at thousands of units by the end of 2026.

Google DeepMind adopted a different Western counterpunch: an expanding partnership strategy across Boston Dynamics, Agile Robots, and Apptronik. Their ambition is to establish the AI foundation-model layer across multiple hardware manufacturers, generating an "Android of robotics" ecosystem. If executed successfully, the hardware race between Chinese manufacturers becomes less commercially decisive because the underlying intelligence layer remains controlled by a Western software stack.

04 — Head-to-Head: China's Giants vs. America's Challengers

05 — Taiwan's Strategic Position: The Island at the Center of the Robot Race

Taiwan holds a position in the humanoid robot supply chain that is both structurally unique and strategically precarious. While Western firms develop the most sophisticated AI models, Asia remains the dominant manufacturing hub for robot components.

Taiwan sits at the apex of this Asian supply chain. TSMC manufactures the AI chips powering nearly every major humanoid platform, from NVIDIA's Jetson Thor to custom silicon developed by Tesla and XPeng. Furthermore, Taiwan's precision manufacturing ecosystem supplies crucial motors, actuators, gearboxes, and sensor assemblies to robotics firms in both the U.S. and China.

For operations focused on AI companion devices and human-machine interaction, the geopolitical friction yields distinct commercial gaps. The massive scale of Chinese hardware lowers component prices globally. The U.S. software superiority produces robust, accessible model ecosystems. The impending consumer humanoid wave will hit Asian markets within 18 months, offering a first-mover window for Taiwanese developers specializing in companion AI tuned for localized cultural nuances.

The global robot race is avoiding a zero-sum conclusion. The organizations that successfully integrate Chinese manufacturing efficiency with U.S. AI intelligence will control the market once it hits scale. Geographically and operationally, Taiwan is suited to execute this exact integration.