PsiBot raises over $100M to scale embodied AI robots

PsiBot’s new nine figure round is another sign that China is pouring serious industrial capital into embodied AI, not just chatbots. The investor mix reads like a who’s who of Chinese manufacturing and local state capital: auto supplier Tuopu, a Chery backed fund, Lens Technology, gaming firm 37 Interactive and multiple municipal vehicles. They are not paying for a demo; they are paying for robots that can actually handle parts, boxes and tools on real factory lines. ([laplaceatlas.com](https://www.laplaceatlas.com/stories/embodied-ai-startup-psibot-raises-over-100-million-in-new-funding-round-57a04125))

In race to AGI terms, embodied AI startups like PsiBot are where model capabilities meet the messiness of the physical world. Their dual system architecture, combining an operation policy model (Psi R2) with an action conditioned world model (Psi W0), is exactly the kind of closed loop needed to turn LLM style reasoning into reliable manipulation. As these systems get better at learning from human teleoperation and high frequency sensor data, they will feed back new training signals that pure text models never see.

Strategically, this funding pushes back against a narrative that China is behind only on frontier language models. If PsiBot and peers can deploy fleets of dexterous arms into logistics and manufacturing, they will generate vast proprietary datasets on real world interaction that could matter as much as extra tokens scraped from the web.