
Robbyant is the embodied-intelligence company under Ant Group, aiming for one brain shared by all robots: a full-stack platform provider built around its in-house cross-embodiment foundation models, rather than a whole-robot vendor.
The operating entity is Shanghai Ant Robbyant Technology Co., Ltd., founded in December 2024 with RMB 100 million in registered capital, wholly owned by Ant Intelligent (Hangzhou) Technology Co., Ltd. and headquartered in Shanghai. On 3 August 2026 the company confirmed the launch of its first external funding round, targeting RMB 1.5 billion.
A full-stack route across models, hardware and solutions. On the model side it open-sources the LingBot family: LingBot-Vision (vision foundation model), LingBot-Depth (spatial perception), LingBot-Map (streaming 3D reconstruction), LingBot-World (world model), LingBot-VLA (cross-embodiment manipulation model), LingBot-VA (video-action world model) and LingBot-Video (video generation). On hardware, wheeled humanoid robots R1, R1-S and R2. On solutions, smart retail, parcel sorting, industrial loading and unloading, and elderly care.
A rare open-ecosystem route in China, where a foundation-model company supplies the brain and robot makers supply the bodies. LingBot-VLA 2.0 has been adapted for robot makers including AGIBOT; a strategic partnership with Leju Robotics covers bodies, data and models, with Leju supplying nearly 10, 000 hours of real-robot data; Robbyant is co-developing an integrated depth camera with Orbbec based on LingBot-Depth.
