Chinese humanoid and quadruped robot maker Unitree made its trading debut on the Shanghai Stock Exchange's STAR Market, with shares surging as much as roughly 500-600% above the IPO price before settling to a smaller, still enormous gain, making it the first publicly listed humanoid robotics company in mainland China. The company raised around $904 million in its IPO, pricing shares at roughly 150.80 yuan (about $22.36) each, and the offering was reportedly oversubscribed by more than 8,000 times, a record for the STAR Market, China's Nasdaq-style board for tech listings. The debut deliberately coincided with the World Robot Conference in Beijing, where hundreds of companies were showing off humanoid robots, robot dogs, and industrial automation systems, underscoring how central robotics has become to China's technology strategy alongside AI and semiconductors. Unitree built its reputation on relatively affordable humanoid and quadruped robots capable of running, dancing, and performing acrobatic movements in viral demo videos, and has increasingly positioned those platforms for industrial tasks like inspection and material handling. For people building software and infrastructure, Unitree's listing is worth tracking less as a robotics story and more as a capital-markets signal: a public listing gives investors, and by extension the broader market, a direct read on how much conviction exists that humanoid robotics can move past demo-stage novelty into profitable industrial deployment, the same transition agentic AI software is being asked to make. If markets keep rewarding robotics and physical-AI companies this aggressively, expect more capital, more competing startups, and faster iteration cycles in the sensors, actuators, on-device compute, and control software robotics companies depend on, several of which overlap directly with edge AI and embedded systems work happening in adjacent software fields. The open question the IPO doesn't answer is whether humanoid robots can actually deliver reliable, profitable performance in real factories and warehouses at scale, rather than in conference demos.