Robbyant and Leju Robot Partner to Accelerate Commercialization of Embodied AI Robotics

Image Courtesy: Public Domain

Robbyant, an embodied AI company within Ant Group, announced a strategic partnership with Leju Robot, a leading company focused on core robotic technologies. This partnership aims to drive innovation centered on embodiment, data, and models of robots, exploring the application and commercialization of embodied AI in real-world scenarios and accelerating the transformation of embodied AI robots from specific task execution to general intelligence.

Under this partnership, the two companies will leverage Leju’s strengths in robotic embodiment, data, and use cases, combined with Robbyant’s expertise in embodied AI models, to cultivate a high-quality dataset of real-robot interactions. Meanwhile, they will join hands to conduct model training, optimization, iteration, and embodiment adaptation to continuously enhance the robot’s full-stack capabilities in perception, understanding, decision-making, execution, and learning. By focusing on industrial and commercial service applications, they also aim to build industry solutions and benchmark use cases, thereby accelerating the commercialization of embodied AI.

Zhu Xing, CEO of Robbyant, said: “The embodied AI industry is evolving from technical verification to real-world deployment. How to seamlessly integrate model capabilities, robotic embodiments, and practical use cases has become a key direction for industry development in the next stage. Through this partnership, we aim to advance the verification and application of related technologies in real-world scenarios and accelerate industrial progress.”

Previously, as a core data partner for Robbyant, Leju provided nearly 10,000 hours of high-quality, multimodal real-robot data for Robbyant’s LingBot-VLA model. LingBot-VLA, a vision-language-action (VLA) model, is designed to serve as a “universal brain” for real-world robotics, helping to reduce post-training costs and accelerate the path to scalable deployment. The model has been successfully adapted to robots from leading manufacturers, demonstrating strong cross-morphology transfer capabilities across diverse robot platforms.

In January, Robbyant released a suite of embodied AI models. Alongside LingBot-VLA, the suite includes

• LingBot-Depth, a high-precision spatial perception model designed to enhance robots’ depth sensing and 3D environmental understanding capabilities in complex real-world environments,
• LingBot-World, a world model that achieves industry-leading performance in video quality, dynamic fidelity, long-term consistency, and interactivity,
• LingBot-VA, an innovative model enabling thinking through autoregressive video prediction.