Unitree’s market debut turned dancing, backflipping robots into a $66 billion equity story almost overnight. Its founder, however, is warning that the technology investors are buying is still nowhere near ready for ordinary homes or factories.

Founded in Hangzhou in 2016, Unitree first built recognition with robot dogs before its humanoids became fixtures of China’s Spring Festival Gala. Their increasingly elaborate performances — including kung fu, jumps and backflips — helped make the company a national-tech champion as Beijing elevated humanoid robotics as a strategic industry. Chinese companies accounted for about 97% of global humanoid-robot shipments in the first half of 2026, with Unitree holding roughly 31%, according to Smart Analytics Global.¹

On Wednesday, that enthusiasm hit the Shanghai STAR Market. Unitree raised about $900 million at a $9 billion IPO valuation, then surged as much as 629% before closing 460% higher. The frenzy briefly put a far richer price on the company than U.S.-based Figure AI, despite Unitree reporting $252 million in 2025 revenue and $41 million in profit. Its IPO valuation was already 219 times 2025 earnings, a measure of how much future growth investors have priced in.¹

A day later, CEO Wang Xingxing supplied the counterweight. He said robotics will have its "ChatGPT moment" only when a machine can enter an unfamiliar home and complete about 80% of tasks from voice or language commands alone.² That could come in two or three years, he said, but perhaps not for five to 10. Today’s robots still need retraining for new tasks and remain less efficient than people.

The split is stark. Nomura sees Unitree’s rapid iteration as a first-mover advantage, while HSBC argues that, without a major jump in AI-model capability, the shipment boom may not last beyond one to two years.³ A new U.S. ban on foreign-made robots also threatens an important overseas market and the customer feedback Unitree uses to improve its products.³

Wang’s caution puts him at odds with Nvidia’s Jensen Huang, who says robots’ breakthrough "happened a couple of years ago already."⁴ For Unitree, the trading debut proved demand for the dream; the harder test is making robots useful enough to earn it.