DexForce

Robotics · Humanoid Robots · AI

Funding

DexForce raises $147M Series B, valuation hits $1.4B for robot training

Simulated training bypasses the costly real-world data collection that bottlenecks most humanoid robot makers today.

• DexForce raised 1 billion yuan (about $147 million) in a Series B round, valuing the Shenzhen startup at more than 10 billion yuan (roughly $1.4 billion), led by Lenovo Capital and Lens Technology.
• Founded in 2021 by Jia Kui, a machine-learning professor at the Chinese University of Hong Kong, Shenzhen, the roughly 500-person company builds AI software that simulates physical environments to train humanoid robots before transferring the learned skills to real machines.
• Jia says a single precise human demonstration can take more than 100 repetitions to capture, which is why DexForce generates synthetic training scenarios in virtual environments instead.
• Jia estimates reaching true 'physical artificial general intelligence' through real-world data collection alone would require 10 trillion samples, which he calls fundamentally inefficient.
• DexForce's simulation software is already used by Toyota, Samsung, Panasonic, Foxconn, Haier and Hisense, while its humanoid robots make coffee and serve customers in malls across more than 15 Chinese cities.
• DexForce is reportedly weighing a Hong Kong listing, though CEO Jia Kui declined to comment on the reports.
• If synthetic training can match the reliability of real-world data, it could sharply cut the cost of scaling humanoid robots industry-wide, but unproven sim-to-real transfer remains the open question.