A San Francisco robotics startup with no revenue is reportedly closing in on a valuation near $20 billion, as investors pour unprecedented sums into companies racing to build the brains behind the next generation of robots. Physical Intelligence is in talks to raise at least $1 billion in a new funding round that would value the two-year-old company at roughly $20 billion, nearly double where it stood just months earlier.
According to a report from Bloomberg, existing backers Avenir and Lux Capital are in talks to lead the round, though terms remain under negotiation and nothing has been finalized. The jump would mark a dramatic climb for a company that was valued at about $11 billion as recently as July 2026, per the same report. As BigGo Finance notes, the proposed terms remain at an early negotiation stage, and a deal of this size would push the valuation center of gravity for the entire embodied intelligence sector even higher if it closes.
No Revenue, No Commercialization Timeline
Despite having no current revenue, Physical Intelligence has become one of the most closely watched names in robotics AI, according to a report from Yahoo Finance Australia. The company opened beta access to its latest robotics models to outside developers this month, allowing them to fine-tune the system for their own uses, the outlet reports. Founded in 2024 by former Google DeepMind researchers alongside academics from Stanford University and UC Berkeley, the company has already raised more than $1 billion to date, according to TechCrunch.
That earlier reporting from TechCrunch quotes a company leader, Lachy Groom, saying flatly, “I don’t give investors answers on commercialization.” The outlet's report also describes Physical Intelligence testing its systems with a small number of partners across logistics, grocery, and even a chocolate maker, but stops short of describing any revenue from those relationships. The company previously raised $400 million at a $2.4 billion post-money valuation in November 2024, a round that drew backing from investors including Jeff Bezos, according to CNBC.
What Pi-Zero Actually Does
At the center of the company's pitch is π0, a general-purpose robot foundation model the company describes as spanning images, text, and actions. Physical Intelligence explains on its own website that the model acquires its capabilities by training on embodied experience collected from real robots, learning to directly output low-level motor commands rather than relying on pre-programmed routines, according to the company's own blog post describing the system.
That approach puts Physical Intelligence squarely inside a much bigger wave of capital flowing into so-called physical AI. Global venture funding to robotics startups totaled $9.9 billion in the third quarter of 2026 alone, more than double the $4.3 billion invested in the same period a year earlier, per data from Crunchbase News. Funding to the robotics sector through the first three quarters of 2026 is up around 185% year over year, with China drawing $13.2 billion in robotics investment this year against $12.1 billion for the United States, the same data shows.
Rivals With Revenue, and Rivals Without
Physical Intelligence is not alone in commanding an eye-popping valuation without much to show for it yet. Figure AI, which builds its own humanoid robots, secured a roughly $39 billion valuation despite deployments limited largely to pilot programs at BMW factories and other sites, according to Business Insider. That same report notes that startups still in research and development are commanding multibillion-dollar valuations, often with little revenue and few deployments to show investors.
By contrast, Skild AI has leaned into a different strategy. The company, which raised $1.4 billion for its Skild Brain foundation model and reached a valuation above $14 billion, has emphasized rapid commercialization and recently crossed $100 million in annual recurring revenue, Crunchbase reports. Germany's Neura Robotics raised a comparable $1.4 billion for its own cognitive robots and supporting AI infrastructure, while FieldAI has already surpassed $100 million in revenue with contracts from more than 30 customers, according to Business Insider's reporting.
Why Robots Are So Hard to Scale
The gap between valuation and revenue across the sector traces back to a fundamental bottleneck: there is no internet-scale corpus of physical-world data the way there is for text or images. As Crunchbase News puts it, the data first has to be collected, and converted from sensor tokens to action tokens, a process that requires robots to repeatedly operate and learn through trial and error in real environments. That interaction data is costly and difficult to scale quickly, and real-world robot performance can be thrown off by pets, lighting changes, or objects a robot hasn't encountered before.
Hardware costs compound the challenge. Robotics companies still face the basic difficulty of manufacturing, maintaining, and deploying expensive physical equipment, Business Insider notes, even as software improves. Investors are increasingly favoring companies that can hit production milestones and land paying customers rather than those still largely in the experimentation phase, according to Crunchbase's analysis of first-half 2026 funding patterns, though exact figures for how much capital flowed into the broader physical-AI category that period vary between reports and were not resolved by either outlet.
A Sector-Wide Capital Surge
Physical Intelligence's talks arrive amid a broader surge touching defense and transportation as well as general-purpose robotics. Defense tech firm Anduril Industries raised $5 billion in May 2026 at a $61 billion valuation in a round co-led by Thrive Capital and Andreessen Horowitz, Crunchbase reports. Autonomous surface vessel maker Saronic raised $1.75 billion that March at a $9.25 billion valuation, while self-driving car company Waymo raised a $16 billion Series D in February at a $126 billion valuation. Andreessen Horowitz itself raised a $1.1 billion fund in August dedicated to the physical infrastructure behind AI, including robotics, according to Business Insider.
Crunchbase's own reporting sums up the uncertainty hanging over the entire category: funding is surging and new unicorns are emerging, but the sector's overall direction remains unresolved. For Physical Intelligence, that tension is explicit — a company with no revenue and no promised commercialization timeline is nonetheless reportedly on the verge of a valuation that would place it among the most richly funded AI startups anywhere, with the final terms of the deal still being worked out.
