Westlake Robotics has closed a nearly 100 million yuan Series A round led by Haiyuan Capital, its fourth raise in six months, taking total funding to 500 million yuan. The money backs development of a unified whole-body large model for humanoids and a talent base for embodied intelligence, pushing toward what the company calls an embodied GPT moment.
The business started in 2024, headquartered in Hangzhou, as the first commercialisation of research from Westlake University’s artificial-intelligence and robotics work. Its route is a full-stack combination of a general brain, a humanoid whole-body cerebellum and a self-developed humanoid body, anchored by two core technologies: the GAE general motion model and a dual pretraining architecture for brain and cerebellum.
GAE delivers millisecond-latency teleoperation, programming-free motion generalisation and cross-body compatibility. Without pre-recorded scripts, a robot replicates any human motion in real time, and one motion-capture rig lets a single operator control multiple robots at once. GAE also outputs a standardised motion-control base for most humanoid bodies on the market, a strong argument for scaled commercial use.
The dual pretraining architecture fuses a vision-language-action model with a world model into a layered collaborative intelligence. Early in 2026 the company launched Westlake o1, a full-stack humanoid, completing the loop from algorithms to pretrained models to hardware. It has signed with the Longyou government to build a county-wide real-world training base and partnered with Alibaba, Intel, JD and the Hong Kong University of Science and Technology.
Westlake Robotics has already won nearly 100 million yuan in orders across research, data capture and power inspection. Looking ahead it plans to serve high-risk industrial inspection, post-disaster search and remote precision assembly, extending human operability into environments too dangerous or remote for people.
Editor’s note: This is an adapted translation of the original Gasgoo report. It has been trimmed and restructured for readability for an international business audience.