宇树宣布：开源VLA大模型！

Core Viewpoint - Yushu Technology has announced the launch of UnifoLM-VLA-0, a visual-language-action model aimed at enhancing physical interaction capabilities in humanoid robots, marking a significant evolution from traditional visual-language models [1]. Group 1: Product Development and Innovations - The UnifoLM-VLA-0 model is designed to evolve from general "image-text understanding" to a "embodied brain" with physical common sense through continued pre-training on robotic operation data [1]. - Yushu's humanoid robots have achieved the highest global shipment volume in recent years, indicating a significant leap in large-scale delivery [8]. - The company has launched a humanoid robot "App Store," allowing users to upload and share trained models for various skills, enhancing user experience and robot capabilities [14]. Group 2: Strategic Partnerships and Market Presence - Yushu has become the robot partner for the Central Radio and Television Station's 2026 Spring Festival Gala, marking its third collaboration with the event [1][3]. - The company has established a national flagship store in Beijing, allowing customers to experience and purchase robot products directly [10][14]. - Yushu Technology has completed its IPO counseling work, indicating its intention to enter the capital market, with established partnerships with several A-share and Hong Kong-listed companies [14]. Group 3: Achievements and Recognition - In 2025, Yushu hosted the world's first humanoid robot combat competition and won four gold medals, showcasing its leadership in the field [5]. - The company has successfully integrated its robots into various performance settings, including theater and concerts, allowing the public to experience the charm of robotics [6].