Ex-Ideal Auto trio launches Zhijian Dynamic, sets record with 100-unit i7 Pro robot delivery

Three former core members of Ideal Auto's smart driving team — Wang Kai, Jia Peng, and Wang Jiajia — have unveiled the results of their first year in robotics. Their startup Zhijian Dynamic announced the delivery of the first 100 units of its i7 Pro full-scenario robot, setting a new industry record for the fastest 100-unit delivery in the embodied AI sector.

Founded in late July 2025, Zhijian Dynamic completed five consecutive funding rounds attracting major financial institutions and two internet giants. The founding trio chaired by Wang Kai with Jia Peng serving as CEO and CTO, and Wang Jiajia as co-founder and COO, all previously held key roles at Ideal Auto's autonomous driving division.

The i7 Pro is a full-scenario industrial robot priced at just 229,800 RMB (about $31,700) for the top configuration — far below the 500,000-600,000 RMB price tags typical of humanoid robots. Wang Jiajia, co-founder and COO, stated that the robot achieves a return on investment within 1.5 years for customers.

More notably, Zhijian Dynamic partnered with Kaixuan Intelligent, a subsidiary of harmonic reducer leader Leaderdrive (绿的谐波), to build what is described as the world's first CNC intelligent embodied robot production line. On this line, i7 Pro robots manufacture harmonic reducer components — essentially the "joints" of robots — creating a hardware self-sufficiency loop where robots build themselves.

According to a report from QbitAI, the journey from lab to factory floor was challenging. During initial testing in real CNC workshops, the i7 Pro struggled with slippery oil-covered floors and imprecise gripper operations on machine controls. Rather than modifying the factory environment, the team chose to improve the robot's adaptability, ultimately creating a general-purpose platform deployable across diverse industrial conditions.

Zhijian Dynamic now employs over 150 people, more than half working on models and algorithms. The company has built the LaST₀ foundation model and post-training system, unifying language, vision, spatial awareness, robot state, and world models into a single Transformer encoding space. Two new products are planned for next month, with expansion targets including retail, logistics, and biopharmaceutical sectors.

Jia Peng draws parallels between the robotics industry today and the early days of electric vehicles, arguing that real-world applications must drive manufacturing maturity rather than waiting for perfect lab conditions.