Xin Lang Cai Jing·2026-02-11 03:32Group 1 - Company completed a 1 billion yuan Series B financing round, with investors including industrial capital, private equity funds, and existing shareholders [1][3] - The company focuses on an end-to-end visual-language-action (VLA) foundational model technology route, continuously enhancing model performance and generalization capabilities [1][3] - The company released the G0 model in August 2025 and its upgraded version G0 Plus in January 2026, claiming to be the "world's first out-of-the-box VLA model" [1][3] Group 2 - The company has achieved thousands of orders, with clients including top global universities, research institutions, and industry leaders [2][4] - It has developed a full-stack self-research capability covering core modules, complete machine design, high-quality data, end-to-end models, and scene solutions [2][4] - The R1 Pro and R1 Lite platforms have become core development platforms for leading laboratories such as Stanford University's Fei-Fei Li team and Physical Intelligence (PI) [2][4] Group 3 - The company has established a high-efficiency data collection and model training foundation, innovating new collection methods like UMI to enhance data collection efficiency, richness, and accuracy [1][3] - In 2026, the company entered a phase of large-scale training with hundreds of thousands of hours of high-quality data, continuously optimizing its embodied foundational model through massive scene data [1][3]