A startup just raised $26 million to make nuclear power plants way smaller and cheaper
Apollo Atomics, a startup from Y Combinator's Spring 2026 batch, has raised $26 million to develop a smaller version of a key part inside nuclear power plants. The company says its design can shrink a reactor by 40 times and cut production costs to a fifth of what they are today. The round, which included $5 million in debt, was led by FCVC with participation from Telesoft Partners, Y Combinator, Alumni Ventures, Robinhood Ventures, Nucleation Capital, Pelion VC, Duke Capital Partners, New Era Ventures, Orange Collective, Stanford University, E14 Fund, and Neutron Power Ventures. The funding was announced exclusively to TechCrunch on August 20, 2026.
The company, co-founded by CEO Assil Halimi and Drew Walker, is not designing a new type of reactor. It is rethinking how heat from a nuclear reactor gets turned into electricity, by focusing on one specific part: the steam generator.
Most nuclear plants today use what is called a pressurized water reactor. In simple terms, the reactor heats water, and that hot water is used to make steam. The steam then spins a turbine, which generates electricity. The steam generator is the component that transfers the heat from the reactor to make that steam. In existing plants, steam generators are enormous — several stories tall — and are built by hand on the construction site, which is one reason nuclear plants take so long to build.
Apollo's version is about the size of a person. It works by running two separate flows of fluid through a compact metal block filled with tiny, needle-thin channels that transfer heat very efficiently. Think of it like a car radiator, but for a nuclear plant: it does the same job as the massive original, just in a much smaller package. Because it can be mass-produced in a factory instead of built piece by piece on-site, Apollo plans to assemble the whole reactor and steam generator at a factory and ship the finished unit to where it is needed.
The design builds on research from MIT, where Apollo has already built a small 40-kilowatt reactor to demonstrate the technology. The company's website cites 15,000+ reactor hours of operating experience and describes its product as "the most power-dense nuclear system." Apollo also plans to offer commercial-grade nuclear fuel and supply chain services alongside its reactor design and operations.
Apollo plans three reactor sizes, ranging from 10 megawatts to 300 megawatts of electricity. For comparison, a typical large nuclear plant today produces about 1,000 megawatts, so even Apollo's biggest reactor is smaller than what most people picture when they think of a nuclear plant. The company expects to build a 300-megawatt power plant in less than 24 months, crediting the faster timeline to less labor and a smaller physical footprint. A demonstration reactor is planned before a commercial deployment targeted for 2028.
The cost target is ambitious. Halimi says an Apollo power plant could generate electricity at 3 cents per kilowatt hour, with the goal of beating natural gas on price rather than just being competitive with existing nuclear plants. Conventional large nuclear plants, as Apollo notes, take more than 10 years and roughly $20 billion per plant to build.
The question that matters is whether building these components in a factory and shipping them really does bring costs down as much as Apollo predicts. The underlying reactor technology is well-understood by regulators, which helps. But the new steam generator is the unproven part — it will need to survive years of repeated heating and cooling, plus radiation exposure, before anyone can confirm the claims. The 3-cent figure would make nuclear cheaper than most natural gas plants, but it is a projection from a company that has not yet built a full-scale unit.
The $26 million raised, even with debt included, is small compared to what nuclear ventures usually need to reach the demonstration stage. Apollo is not trying to invent a new reactor core, which makes its approach cheaper to develop. But there is still a big gap between this seed funding and a 2028 commercial deployment, with a demonstration reactor still to come.
What Apollo is proposing is narrower and potentially easier to pull off than designing an entirely new type of reactor. If the compact steam generator works as described, it could change how nuclear plants are built and where they can be placed. The range of sizes on offer, from 10 to 300 megawatts, could serve both small local power needs and larger utility-scale projects. A construction timeline under 24 months would be a dramatic improvement over how long nuclear plants take today. The technology is at the demonstration stage now, and the 2028 commercial target will be the number to watch.
