Alibaba Group has launched RynnBrain, an open-source embodied AI model designed to enable robots and smart devices to perform complex real-world tasks. The model, developed by Alibaba's internal research division DAMO Academy, was released on Tuesday and is now available on GitHub and Hugging Face.
RynnBrain is trained on Alibaba's Qwen3-VL vision-language technology and comes in multiple versions, starting with a 2-billion-parameter model and including a more efficient mixture-of-experts configuration. The AI system allows machines to recognize objects, understand spatial relationships, predict object trajectories, avoid obstacles, and plan sequential actions in environments such as factory assembly lines and kitchens.
Alibaba claims the model outperforms competing systems from U.S. tech giants in benchmark tests, specifically naming Google's Gemini Robotics-ER 1.5 and Nvidia's Cosmos-Reason2. This launch positions Alibaba in direct competition with these companies in the rapidly growing field of physical AI and robotics.
The release is part of a broader strategic push by China to dominate physical AI and robotics technology, contrasting with the more proprietary approach often taken by U.S. firms. Jeff Zhang, Alibaba's Chief Technology Officer and head of DAMO Academy, stated: "The DAMO Academy will be at the front of developing next-generation technology that supports Alibaba and our partners. We want to discover new tech that helps users and businesses work faster and more securely."
Alibaba is expanding its global research footprint with seven new labs in Beijing, Hangzhou, San Mateo, Bellevue, Moscow, Tel Aviv, and Singapore. The company also plans to recruit 100 researchers worldwide specializing in IoT, data intelligence, and natural language processing. DAMO Academy is collaborating with academic institutions, including a partnership with the University of California, Berkeley's RISE Lab on secure real-time computing.
Following the announcement, Alibaba's stock (BABA) rose 0.30% in early trading. Wall Street analysts maintain a Strong Buy consensus rating on the stock, with an average price target of $203.09, implying a 24.60% upside potential.
