Jeff Bezos has stepped to the center of one of the biggest wagers yet on artificial intelligence applied to the physical world, with his startup Prometheus raising $12 billion in a Series B round at a valuation of roughly $41 billion. The company announced the round on June 11, 2026, and it ranks among the largest single funding events ever tied to what the industry calls physical AI. Prometheus is co led by Bezos and Vik Bajaj, a Stanford scientist who helped found the health venture Verily, and the two have framed the company around a single ambitious idea.
That idea is an artificial general engineer. Where most of the attention in artificial intelligence has gone to systems that produce text, images, and code, Prometheus is building software meant to design and manufacture complex physical systems. The company points to targets such as jet engines and drug compounds, products that today require years of specialized human work, expensive prototyping, and slow cycles of testing. The bet is that the next major leap in artificial intelligence will be measured not in words generated but in machines and molecules designed, a shift from software that talks to software that builds.
The money behind the round reflects how seriously large investors are taking that thesis. The Series B drew JPMorgan, BlackRock, Goldman Sachs, DST Global, and Arch Venture Partners, with Bezos himself among the backers. Counting earlier funding, Prometheus has now raised about $18 billion, a war chest that puts it among the best capitalized startups in the field before it has shipped a public product. The company runs with roughly 150 employees, a small head count for the size of the raise, which signals a strategy built around heavy compute and top tier talent rather than a large work force.
Bezos has pushed back on the secrecy that surrounded the company in its early months. Speaking after the round, he said the company is not being secretive about its aims, an attempt to reframe a startup that had operated quietly into one with a clear public mission. The openness matters because the promise is enormous and unproven. Designing a jet engine or a new drug involves physics, materials science, safety testing, and regulation that text based models never touch, and turning artificial intelligence loose on those problems requires both new methods and new ways to verify that the output actually works in the real world.
The appeal for investors lies in the size of the markets Prometheus is aiming at. Aerospace, pharmaceuticals, advanced manufacturing, and energy are vast industries where even modest gains in design speed or efficiency translate into huge value. A system that could shorten the path from concept to working hardware, or from molecule to validated drug candidate, would touch parts of the economy that generative AI has barely reached. That potential is what justifies a $41 billion valuation for a company still early in its work, since the prize is a foothold in the physical economy rather than another chatbot.
The challenges are just as large as the opportunity. Physical AI has to contend with the messiness of the real world, where a design that looks elegant in simulation can fail under heat, stress, or manufacturing limits that software does not capture. Prometheus will need to prove that its artificial general engineer can produce results that engineers and regulators trust, and it will have to do so in fields where mistakes carry safety consequences that a wrong sentence never does. The company also enters a crowded race, with other firms and established industrial players exploring how to fold artificial intelligence into design and manufacturing.
What Prometheus represents is a widening of the artificial intelligence story beyond screens and servers. Bezos is betting that the same kind of capability now reshaping software can be turned toward the hard problems of building real things, and he has assembled the capital and the talent to test that idea at scale. Whether the artificial general engineer becomes a working tool or an expensive ambition will take years to judge, but the $12 billion round makes clear that some of the largest investors in the world are willing to fund the attempt.