FieldAI's Robot-Brain Software Is Reportedly Worth $10 Billion. Nobody Will Say Who's Paying For It.

FieldAI, an Irvine, California startup that builds autonomy software for humanoid robots, robot dogs, drones, and industrial rovers, is in the middle of a financing that values the company at roughly $10 billion — about five times the $2 billion mark it carried a little over a year ago. Reports on the deal first surfaced October 2 and were still being corrected and re-reported as of October 6, a week in which outlets could not agree on the most basic fact about the raise: whether it has actually closed.

A deal that keeps changing shape mid-flight

The initial report, credited to Business Insider reporter Rya Jetha, described a signed term sheet for $700 million — language that implies negotiations were still active rather than final. Within days, at least one outlet was framing the same raise as a closed Series C, while a Benzinga write-up published four days later corrected its own earlier framing to clarify that FieldAI was still only "planning" to raise the money. No version of the story names a lead investor, and FieldAI itself has not confirmed the raise to any outlet that asked. That a $10 billion valuation can circulate for a week without anyone — company included — confirming who is actually writing the check is itself notable for a sector investors are currently pricing at generational multiples.

We have seen very, very fast growth in the last several months.

That line, attributed to FieldAI chief executive Ali Agha in reporting from June, is the only public comment from the company that has surfaced across the coverage of this round — and it predates the $10 billion figure by several months, meaning it was made about the company's trajectory generally, not about this specific valuation.

The multiple behind the number

The clearest way to stress-test the valuation is to look at what FieldAI says it is selling against what it says it is worth. Reporting consistent across multiple outlets puts the company's combined revenue and signed customer contracts at just over $135 million, spread across more than 30 customers in construction, data centers, and defense — with roughly $35 million of that added since June. None of the reporting distinguishes recognized, recurring revenue from the face value of signed contracts, a distinction that matters enormously for a company being priced like an infrastructure platform. Taking the $135 million figure at face value anyway, a $10 billion valuation implies a multiple above 70 times that combined metric. For comparison, even the richest software-as-a-service multiples during the 2021 financing peak rarely cleared 30 to 40 times forward recurring revenue, and FieldAI's figure isn't purely recurring revenue to begin with.

Priced against a crowded field, not a monopoly

FieldAI is not raising in a vacuum. Physical Intelligence has reportedly been in talks at a valuation above $11 billion, and Skild AI raised at a valuation above $14 billion earlier this year — both companies pursuing a similar thesis that general-purpose "robot brain" software, rather than any single hardware platform, is the layer worth owning as humanoids and other autonomous machines move from demos to deployment. FieldAI's pitch leans on breadth: its software is reported to run across quadrupeds, humanoids, wheeled robots, and passenger-scale vehicles, and the company has reportedly pulled engineering talent from Google DeepMind, Tesla, Nvidia, and Boston Dynamics. But three well-capitalized competitors chasing the same software layer, each backed by marquee venture names, means investors are underwriting a race with no established winner — and a round this size bets heavily that FieldAI's 30-plus customer relationships translate into durable lock-in before a rival's platform does the same.

The money behind FieldAI's prior $405 million, raised across two tranches roughly a year ago at the $2 billion valuation, reads like a checklist of strategic and sovereign capital rather than pure financial investors: Nvidia's venture arm NVentures, Intel Capital, Jeff Bezos' family office, Bill Gates' Gates Frontier, Samsung, Khosla Ventures, Singapore's Temasek, and Emerson Collective, the Laurene Powell Jobs-backed fund. That roster suggests chipmakers and sovereign-adjacent funds are positioning across the physical-AI stack the way they positioned across large language models — a pattern that raises the stakes if the robotics deployment timeline runs slower than the financing timeline suggests.

What it means for investors

The through-line connecting FieldAI's round to the Physical Intelligence and Skild AI valuations is a wager that humanoid and field-robot software is close to an inflection point industry analysts have flagged as still largely unproven — most humanoid deployments remain pilots or teleoperated demonstrations rather than fully autonomous production use, with locomotion generally further along than dexterity. Nvidia CEO Jensen Huang has talked up a total addressable market in the tens of trillions of dollars for humanoid robots and physical AI broadly, a figure that has become a standard reference point for valuations in this category even though it describes a potential future market, not current revenue. For investors watching the Markets & Financials space, FieldAI's round is a useful marker of how far late-stage robotics valuations have detached from disclosed financial metrics: a legitimate, well-connected startup can apparently command a 70-times-plus multiple on a revenue-and-contracts figure that isn't fully audited, in a financing round whose lead investor and closing status remain unconfirmed a week after it became public. Whether that gap closes through FieldAI actually scaling deployments, or through a valuation correction once the deal's real terms surface, is the open question this round leaves for anyone pricing the broader physical-AI trade.