Panhǔ Intelligent Completes Angel Round of Tens of Millions of Yuan, With Puhua Capital and Hengze Capital as Co-Investors

Hangzhou-based embodied intelligence startup Panhǔ Intelligent was reported to have completed an angel round of tens of millions of yuan on October 20, 2026. The investors were Puhua Capital and Hengze Capital. The company said it will use the funds to develop the TigerMind multidimensional sensory cognition world model, edge intelligence and industrial-grade embodied intelligence systems, and to conduct joint validation with industrial partners in real-world scenarios.

Panhǔ Intelligent’s approach combines a perception layer integrating vision, force and touch with joint validation in real-world scenarios. This directly addresses challenges in embodied intelligence, where reproducibility and practicality in the field matter alongside accuracy in the lab. The investment by Puhua Capital and Hengze Capital shows that the investors have put funding into these underlying technologies.

The focus of this round is not just that Panhǔ Intelligent secured tens of millions of yuan in angel funding, but that the funds are earmarked for a world model, edge intelligence and industrial-grade embodied intelligence systems. The money is intended not simply for research and development, but to establish a technical framework combining multidimensional sensing across vision, force and touch with the understanding and prediction of physical states. This suggests a focus on building a foundation for real-world validation rather than simply developing a demonstration robot. A related article from this newspaper, Gewu Tongzhi Raises Nearly 100 Million Yuan in Angel Funding to Develop a “Brain” That Understands Physical Changes, showed that understanding the physical world has become an investment theme in embodied intelligence. Panhǔ Intelligent’s use of the funding follows that trend. One difference is that the company puts a world model at its core while also focusing on edge processing and joint validation in industrial settings. The challenge is not only the performance of a trained model, but whether it can be implemented in a system that operates in the field. In terms of industry structure, this funding goes to perception, inference and control layers upstream of competition over robot hardware. Hou Dibo’s research background and Lin Feng’s work experience at Alibaba, Ant Group and DingTalk appear to inform an approach designed to connect research-driven technology with industrial adoption. However, what has been disclosed so far is limited to the scale of the funding, the investors, its intended uses and the areas of focus. The company has not disclosed production volumes, commercialization timing, deployment sites, how training data will be collected or evaluation metrics. The next questions are what types of real-world scenarios the world model can prove effective in, and how much responsibility it will take on for edge processing and on-site control.