Realtime AI News
Physical Intelligence in talks to raise $1B at a reported $20B valuation, with no revenue
Physical Intelligence is negotiating a new round of at least $1 billion that would value the robot foundation-model company at about $20 billion, Bloomberg reported. The San Francisco startup, founded in 2024, has no commercial revenue and builds no robots of its own, betting instead on a general robot brain that runs on third-party hardware.
Physical Intelligence, the San Francisco robot AI company, is in talks to raise at least $1 billion in a new round that would value it at about $20 billion, according to Bloomberg, which cited people familiar with the matter. Existing investors Avenir and Lux Capital are considering leading the round.
If the deal closes, it would be another outsized bet on embodied AI. Physical Intelligence raised $600 million at a $5.6 billion valuation last November, and its valuation had risen to $11.5 billion earlier this year, so the figure now under discussion is nearly double that.
What stands out is the contrast with revenue: the company has no commercial revenue and no mass-produced robot of its own. QbitAI, which reported the talks, described it as a company worth $20 billion without earning a dollar.
The investor list explains part of the premium. Beyond Avenir and Lux Capital, Alphabet's CapitalG, Thrive Capital and Sequoia are shareholders, and Jeff Bezos bet early in a personal capacity.
The team explains the rest. Founded in 2024, Physical Intelligence is led by CEO Karol Hausman, a longtime Google and Google DeepMind robotics researcher, alongside co-founders Sergey Levine, a UC Berkeley professor, Chelsea Finn, a Stanford assistant professor, and Lachy Groom, an early Stripe employee turned angel investor.
Its approach is pure software. Rather than building hardware, it adds action generation to vision-language models to train vision-language-action (VLA) models that can drive third-party robot bodies, with the line running from π0 and π0.5 to this year's π0.7.
π0.7, released in April, showed progress on compositional generalization: a two-armed robot never trained specifically to fold laundry completed the task using knowledge learned from other robots, even though the arms differed in size and structure. This month the company also opened a beta API for its robot foundation model, offering fine-tuning and inference to outside developers.
Rivals show how aggressive the pricing is. Skild AI, which also builds a general robot brain, was valued at $14 billion in January and said in September that annualized recurring revenue had passed $100 million; Figure AI is valued at $390 billion.
The company itself flags the risk. In opening its beta, Physical Intelligence warned that real-world robotics is genuinely hard, and that construction, robot surgery and warehouse automation each demand the resources of an entire company. Whether the round closes and whether the API draws real customers are the next things to watch.
Why it matters
If the round closes, Physical Intelligence becomes one of the most valuable embodied-AI companies before shipping a commercial product, widening the gap between sector valuations and actual revenue.
Nearby Updates
All10/28, 16:57
Doubao denies Hengdian extras data fed ByteDance models
Doubao public relations lead Liu Xing said on October 28 that claims Hengdian and ByteDance fed the portraits and voices of 150,000 film extras into a ByteDance model are seriously inaccurate. He said the Hengdian film AI platform exists to build high-quality datasets for productions, that its data belongs to Hengdian or its licensors, and that it will not train ByteDance models.
10/28, 16:59
MoleculeMind reports 18 of 19 design projects hit, with all six GPCR targets bound
MoleculeMind published a technical report on MMDesign, its AI protein design platform, covering 19 design projects across 18 targets, 18 of which produced experimentally confirmed binders for a 94.7% project success rate. The report also claims its in-house MMFold structure model roughly doubles AlphaFold 3's top-1 high-quality success rate on antibody-antigen targets in the FoldBench benchmark.
10/28, 17:04
Rokid adds a full-color AI glasses model starting at 4,999 yuan, with a ring for input
Rokid has introduced a new full-color AI glasses model starting at 4,999 yuan, adding a ring for gesture input, according to a retail roundup published by Ebrun on October 28. The report frames the launch as part of an escalating race in AI glasses but gives no specifications, ship date or battery detail.
10/28, 17:30
Legora builds an agentic operating system for legal work on OpenAI models
OpenAI's October 28 customer story details how Stockholm-based Legora uses its models to power contract review, due diligence and agentic legal workflows. The case study says NDA review dropped from about 20% of Legora's legal-team capacity to under 1%, while one Fortune 100 customer cut contract turnaround by 90%.