Ant Lingbo Showcases Full-Stack Open-Source General-Purpose Robot Brain at Wuzhen

From November 2 to 5, the 2026 World Internet Conference Wuzhen Summit was held in Wuzhen, Zhejiang, under the theme "Open Source and Openness, Joint Contribution and Shared Benefits—Working Together to Build a Community with a Shared Future in Cyberspace." Ant Lingbo, the embodied intelligence company under Ant Group, brought its fully open-source general-purpose robot brain to Wuzhen, showcasing new progress in model development, open-source collaboration, and scenario applications.

Today, open source is extending from language, code, and content generation to robots' perception, prediction, and action in the physical world. Unlike the digital world, robot bodies vary greatly and task scenarios are highly dispersed, forcing many teams to train a "brain" from scratch for every configuration and every scenario.

In response, Lingbo's approach is an open embodied foundation model. Lingbo has now open-sourced to the world its core models covering spatial intelligence, dexterous manipulation, and environmental feedback, building full-stack capabilities around making robots "see more clearly, think more intelligently, and work more efficiently," so that the same set of intelligence can be reused across different robot configurations, tasks, and scenarios.

Ant Lingbo Brain Full-Stack 2.0 Model Matrix

According to statistics, in the nine months since the LingBot series models were open-sourced, downloads have exceeded 2.2 million, GitHub Stars have surpassed 33,000, and the related work has been cited by global peers such as Physical Intelligence and Generalist AI. This robot brain from China has entered the global research and engineering ecosystem: Stanford University's OpenWAM project reused and adapted the LingBot-VA module; South Korea's FastOPD research used LingBot-VLA as a teacher model to explore lighter-weight robot training solutions; and the Swiss Federal Institute of Technology Lausanne (EPFL) conducted three-dimensional coral reef reconstruction based on LingBot-Map.

At the same time, Lingbo models are accelerating their integration into global developer toolchains: LingBot-VLA and LingBot-VA have completed ecosystem adaptation for LeRobot, the world's largest robot development platform; LingBot-Vision has been merged into the mainline of the industry's mainstream vision backbone model library timm; LingBot-World has become the world's first world model to simultaneously gain support from the two major mainstream open-source inference engines vLLM and SGLang, and NVIDIA has also integrated LingBot-World into its open-source world model inference library FlashDreams.

The vitality of the open-source ecosystem is also reflected in the developer community. The online preliminary round of the first "Ant Lingbo Embodied Large Model Challenge" recently concluded, attracting 1,855 teams and 3,252 contestants from 648 universities and enterprises in 13 countries, setting a record for the largest scale in the industry.

Ant Lingbo's general-purpose brain drives a robot to autonomously retrieve medicine

During the summit, Ant Lingbo and Guoda Pharmacy jointly presented a robotic smart pharmacy at the Wuzhen World Internet Science and Technology Museum. The booth fully recreated a real pharmacy environment, with medicines densely arranged; in an 80-centimeter-wide narrow aisle, the robot autonomously completed the entire continuous workflow from order reception and intelligent sorting to product delivery. Currently, the related solution has been operating routinely in multiple Shanghai Guoda Pharmacy stores; without modifying product displays or store layouts, the robot can integrate into the existing operating environment and help pharmacists share the high-frequency, repetitive sorting work.

Beyond pharmacies, Lingbo is working with partners such as Longsheng Weirui, Dema Technology, and De'er Xiangwang to expand industrial loading and unloading and logistics sorting scenarios, validating "one brain, multiple machines" with robot body manufacturers such as Leju and Tihu, and has signed a memorandum of cooperation with the Arab Federation for Digital Economy (AFDE) to promote the application of the general-purpose robot brain in industries such as retail, logistics, and hotels and catering in the Middle East.

With technological innovation and practical results, Ant Lingbo won two honors in the "New Light" series of the 2026 World Internet Conference "Light of the Internet" Expo: the robot general-purpose brain full-stack model won the "New Light" Product Award, and the robotic smart pharmacy won the "New Shine" Scenario Award.

Leiphone Leiphone