On April 7, 2026, Qianxun Intelligent Technology announced the completion of a new financing round, raising 1 billion yuan. This round was co-led by Shunwei Capital and Yunfeng Capital, with significant contributions from Datian Caizhi, a major RMB fund, along with Yinhe Yuanhui, Turing Fund, Xinding Capital, and Gengxin Capital. This latest financing comes just a month after Qianxun Intelligent successfully completed two rounds totaling nearly 2 billion yuan at the end of February, with investors including Yunfeng Capital, Chaos Investment (led by Ge Weidong), Sequoia China, Synstellation Capital, and TCL Capital. Following this round of funding, the company’s valuation has surpassed 10 billion yuan.
In January, Qianxun Intelligent released its open-source Spirit v1.5 model. In the RoboChallenge Table30 rankings, Spirit v1.5 achieved a score of 66.09 and a success rate of 50.33%, surpassing PI0.5 and becoming the first Chinese open-source model to exceed its performance.
The company is advancing its data system construction with a focus on diverse data. It has accumulated over 200,000 hours of various real interaction data, covering dimensions such as internet videos, remote operation, and wearable data collection. The total data volume is expected to exceed 1 million hours by 2026. Additionally, Qianxun Intelligent has developed its fifth-generation wearable data collection device, significantly reducing collection costs to one-tenth of traditional methods, thus eliminating cost barriers for large-scale data collection. As of April 2026, the company plans to expand its data team to 1,000 members, continuously providing high-quality real physical interaction data for model iteration.
In terms of hardware, Qianxun Intelligent previously launched the full-force humanoid robot Moz1. As the first high-precision, fully-body force-controlled humanoid robot in China, Moz1 is equipped with 26 degrees of freedom and features the world’s highest power density integrated force control joints, along with a high-speed WBC algorithm, achieving a load-to-weight ratio of 1:1. This robot can perform zero-delay remote operations, and together with its self-developed multidimensional data collection devices, it can efficiently complete data collection to support hourly model iterations.
On the commercialization front, on March 19, the company officially signed a strategic cooperation agreement with JD.com. The self-developed Moz robot has been fully integrated into the JD.com MALL smart retail scenario. In high-precision coffee-making tasks, the Moz robot not only achieved stable operation in real-life settings but also successfully demonstrated a closed-loop “data collection and model iteration” cycle in a genuine commercial environment. Previously, the world’s first humanoid embodied intelligence production line was launched at CATL‘s Zhongzhou base, with the Mo robot becoming the core equipment of the line, achieving near-zero defect mass production of nearly 1,000 batteries at a production rate comparable to or even faster than skilled workers.
The core team of Qianxun Intelligent comes from prestigious institutions such as UC Berkeley, Carnegie Mellon University, Tsinghua University, and Peking University, as well as companies like ByteDance, Xiaomi, and Tencent. This all-star team boasts cutting-edge AI technology capabilities, experience in robotic hardware engineering, and a vision for commercialization, with an average age of under 30.
Founder and CEO Han Fengtao has over a decade of experience in the robotics industry and was previously the co-founder and CTO of Rocky Robotics, a leader in high-performance lightweight industrial robots. He has led his team to successfully deliver over 20,000 units of various product models. Under his leadership, the team has also implemented commercial success across more than 20 industries and hundreds of scenarios, serving thousands of clients.
Co-founder Gao Yang studied computer science at Tsinghua University and later pursued a Ph.D. in computer vision at UC Berkeley, under the mentorship of renowned expert Pieter Abbeel in the field of embodied intelligence. His research focuses on cross-modal interaction in robotics, with numerous academic contributions recognized at top conferences like NeurIPS and ICRA. Additionally, he serves as an assistant professor at Tsinghua University’s School of Interdisciplinary Information and is the director of the university’s Vision and Embodied Intelligence Laboratory, supervising Ph.D. students and bringing ten years of research experience in embodied intelligence, machine vision, and machine learning.
Original article by NenPower, If reposted, please credit the source: https://nenpower.com/blog/robotic-unicorn-backed-by-jack-ma-and-lei-jun-secures-additional-1-billion-in-funding/
