Back to Search
Physical Intelligence

Physical Intelligence

Companyactive

Physical Intelligence (π) is an AI and robotics company founded in 2024, based in San Francisco, California, USA. The company's mission is to bring general-purpose AI into the physical world by developing learning algorithms that create models capable of controlling any robot to do any task. Physical Intelligence was launched in November 2024, raising $400M at a $2B valuation in its founding round. Key investors included Bond, Jeff Bezos, Khosla Ventures, Lux Capital, OpenAI, Redpoint Ventures, Sequoia Capital, CapitalG (Alphabet), and Thrive Capital. The company's flagship model is π0 (pi-zero), a vision-language-action (VLA) foundation model released on October 31, 2024. π0 is trained on diverse multi-robot data and can output motor commands at up to 50 Hz via flow matching. It was the first generalist robot policy capable of controlling multiple robot types and performing dexterous tasks like folding laundry, bussing tables, bagging groceries, and packing items. Subsequent model releases include π0.5 (April 2025) with open-world generalization for mobile manipulators, π*0.6 (November 2025) with RL-based improvement, π0.7 (April 2026) with emergent capabilities and steerability, and MEM — a multi-scale embodied memory system (March 2026) enabling tasks longer than ten minutes. In February 2026, the company published "The Physical Intelligence Layer," describing how its partners were solving real-world problems with general-purpose physical intelligence models. The company maintains a strong research focus and has open-sourced π0 weights and code.

Details

Updated:6/8/2026
founded year2024
headquartersSan Francisco, California, USA
total funding usd400000000.00

Tags

roboticsVLAfoundation-modelAIgeneralistpolicy

Relationships

Sources

https://www.pi.website
website
Visit
https://www.pi.website/blog/pi0
website
Visit
https://www.nytimes.com/2024/11/04/business/dealbook/physical-intelligence-robot-ai.html
website
Visit

Related Knowledge Pages

No related knowledge pages.