Prometheus reportedly closed a $12 billion Series B at a valuation of approximately $41 billion, according to GeekWire, citing Axios. The company, co-led by Jeff Bezos and Vik Bajaj, describes its mission as building what Bezos and Bajaj call an “artificial general engineer”, an AI system capable of designing real-world physical objects and automating physical product development. That’s a meaningfully different target than the software-layer AI systems most of the market has been funding.
Investors reportedly include JPMorgan, BlackRock, Goldman Sachs, DST Global, and Arch Venture Partners, according to Axios. An independent report from CrainCurrency separately confirmed DST Global and Arch Venture Partners’ participation, providing partial corroboration. JPMorgan, BlackRock, and Goldman Sachs participation traces to Axios reporting only and has not been independently verified. All figures should be read as reported, not confirmed, no SEC Form D filing or official press release was available at time of publication.
The round’s scale matters for two reasons. First, the dollar amount. $12 billion is roughly what Anthropic raised across its first three major rounds combined. That kind of capital commitment, at this stage, signals that institutional investors aren’t hedging on physical-world AI, they’re treating it as a distinct category with a timeline. Second, the investor composition. JPMorgan, BlackRock, and Goldman Sachs aren’t typical early-stage venture participants. Their reported presence suggests this isn’t purely a return-maximization play. Industrial AI that touches manufacturing, physical design, and supply chain infrastructure is exactly where those institutions have client exposure at scale.
The real story isn’t the valuation. It’s the category claim.
Prometheus isn’t positioning itself against OpenAI or Anthropic. The “artificial general engineer” framing targets sectors, automotive, aerospace, industrial manufacturing, advanced materials, that the software AI wave has largely bypassed. Those industries run on physical constraints that language models can’t simulate. Prometheus is betting it can close that gap. That’s a vendor characterization, not an independently verified capability. But the capital behind it suggests institutional investors find the thesis credible enough to act on at scale.
This is also the third major capital event targeting physical-world AI in 72 hours. PhysicsX closed a $300 million round at a $2.4 billion valuation on June 8, targeting the simulation and physics modeling layer beneath AI data centers. The pattern is legible: capital is moving toward physical-world AI applications at a pace that wasn’t visible six months ago.
What to watch
A Form D filing with the SEC would be the first T1 verification of round size and investor participation. Watch SEC EDGAR for a Prometheus filing in the coming weeks, that’s the document that converts “reportedly” into confirmed. Also watch whether the $41 billion valuation holds in secondary markets, which tend to price these rounds more conservatively once the announcement cycle fades.
The catch is that “artificial general engineer” is a positioning term, not a product milestone. Prometheus hasn’t shipped. The capital commitment is real; the capability is a thesis. Investors and corporate strategists evaluating industrial AI exposure should track the Form D first, then watch for any pilot partnerships with named manufacturing or engineering clients, that’s where the thesis either gets traction or stalls.