PsiBot hits US$1.48B valuation with new US$100M round for embodied AI
What's the deal? Chinese AI startup PsiBot is raising close to US$100 million at a US$1.48 billion valuation, becoming a new unicorn in embodied AI and world models. The round is led by carmaker Chery AutomobileDealroom has a profile for this one. Try Dealroom →, with participation from Lens TechnologyDealroom has a profile for this one. Try Dealroom →, a sensor maker for Apple and Tesla.
The money math: Also known as Lingchu Intelligence, the company has raised about US$300 million since its founding in 2024. It was established in Shanghai by Viktor Wang, a PhD from George Washington University; Xiaojie Chai, a robotics veteran of Alibaba and Tencent; Yaodong Yang, an assistant dean at Peking University; and Yuanpei Chen, a Stanford visiting scholar advised by AI scientist Fei-Fei Li.
What's the endgame? PsiBot builds world models that help robots and self-driving cars see and respond to their physical surroundings. Its models learn from real-world data gathered with bespoke hardware such as gloves and humanoid robots, and it is testing platforms at a large Chinese logistics vendor and one of the world's largest fibre-optic cable makers.
Why now? Embodied AI and world models are emerging as the next frontier in artificial intelligence and a key battleground in the US-China race. "World models aim to achieve something more consequential than the large-language models and chatbots," Wang said.
What could go wrong? Wang said data is China's clear advantage but also its primary bottleneck, citing high collection costs, scarcity, lack of standardisation and poor quality. "Even the frontier AI labs in Silicon Valley such as OpenAI and Meta don't have good data," he said. PsiBot plans to collect one million hours of data in 2026 to build a generalised world action model.
The signal: Chinese companies are among the world's most aggressive builders of world models, backed by state support, abundant industrial data and an open-source ecosystem. Wang expects major progress within 24 months, comparing it to a turning point in text AI: "We should see the GPT-2 moment in embodied AI in two years."
Read more: Bloomberg · The Business Times · The Straits Times
