A 20-megawatt reactor that arrives on a flatbed, plugs into a factory, and bills you for electricity - not for a decade-long construction saga.
Somewhere outside Bryan, Texas, a nuclear power plant is being assembled the way a car is assembled - from parts that showed up on trucks. No cooling tower the size of a stadium. No river. No fifteen-year permitting odyssey. Just a slab of concrete and a steel-encased reactor that Last Energy calls the PWR-20, scaled down here to a 5-megawatt pilot at the Texas A&M-RELLIS campus.
This is Last Energy in mid-2026: a 67-person company with roughly $164 million in the bank, a UK site working through regulators, and a fleet of power deals signed across two continents. It does not, technically, sell reactors. It sells electrons. The reactor is just how the electrons get made.
Nuclear power. Radically simplified.- Last Energy, company tagline
The trouble with nuclear power is not the physics. The physics has been settled since the 1950s. The trouble is the project. Conventional plants are gigawatt-scale, custom-built one-off megaprojects, financed by governments and utilities, and they run famously over budget and over schedule. The industry's instinct, for seventy years, was to build bigger. Bigger reactors, bigger price tags, bigger delays.
Meanwhile, the customers who actually need clean, steady power - data centers, steel mills, chemical plants - cannot wait a decade or write a multi-billion-dollar check for a plant they would then have to operate. Renewables are cheap but intermittent. The grid is congested. And demand, thanks largely to AI data centers, is climbing faster than anyone planned for.
Bret Kugelmass did not start in nuclear. He started a drone company, sold it, then spent a few years interviewing nuclear experts for a podcast - the Titans of Nuclear - under a research nonprofit he founded called the Energy Impact Center. The conclusion he reached was contrarian and a little impolite: the nuclear industry's obsession with scale was the bug, not the feature.
So in 2019 he spun out Last Energy with a bet that sounds almost too simple. Make the plant small - 20 megawatts, not 1,000. Standardize it so every unit is identical. Build it in a factory and ship the modules on trucks. Then borrow the discipline that lets the oil and gas industry build enormous, repeatable infrastructure on schedule, and apply it to the atom. Don't ask governments to pay. Use private capital, and sell the output.
Build, own, and operate power plants for customers - reliable on-site power without requiring customers to develop, finance, or operate the asset.- Last Energy, on its business model
The PWR-20 is a single-loop pressurized water reactor - the most boring, proven, and well-understood reactor design on earth, which is exactly the point. Last Energy is not inventing new physics. It is repackaging old physics into something you can manufacture and ship. The whole plant fits inside roughly a football field. It is air-cooled, so unlike almost every other water reactor, it does not need to sit beside a river or coastline.
For pilots and demonstrations, Last Energy scales the design down to a 5-megawatt PWR-5 - the version headed to Texas A&M to prove low-power criticality first, then grid electricity later. Same idea, smaller package.
Bret Kugelmass founds the Energy Impact Center, a clean-energy research nonprofit - and later the Titans of Nuclear podcast.
Last Energy spins out as the commercial venture, built around small, standardized, factory-made reactors.
Series A: $20M led by Gigafund. The open-source OPEN100 reactor design is also released.
Power deals and agreements signed across Poland, Romania and the UK - press reports value the pipeline at up to $18.9B for 34 reactors.
Series B: $40M, with Gigafund, the Autodesk Foundation and family offices. Total raised passes $64M.
Enters UK nuclear site licensing - the first new commercial site since 1978 - and accepts a National Grid connection for its South Wales unit.
Announces first US deployment: a PWR-5 pilot with the Texas A&M System at RELLIS, targeting criticality by mid-2026.
Series C: $100M oversubscribed, led by the Astera Institute. Total funding reaches roughly $164M.
Ambition is cheap in nuclear. Plenty of startups have renderings. Last Energy's harder-to-fake evidence is the pipeline: power purchase agreements for 34 PWR-20 units with industrial partners in Poland and the UK, a 10-reactor plan for Poland's Legnica Special Economic Zone, a cooperation deal with Enea Group, and a four-unit project on the bones of a former coal plant in South Wales.
Then there is the regulatory grind, which is where nuclear dreams usually go to die. Last Energy entered the UK site licensing process and cleared a Preliminary Design Review with the Office for Nuclear Regulation. The U.S. Export-Import Bank issued a letter of interest for up to $103.7 million. None of that is a working plant yet - but it is the unglamorous paperwork that working plants are made of.
Roughly half of its commercial agreements target data centers - the customers most desperate for power that's clean and never blinks.- on why nuclear's smallest reactor found its biggest market
Strip away the engineering and the pitch is about time and access. Last Energy's stated mission is to cut the cost and construction time of nuclear so that clean, reliable power can be deployed quickly, and so industrial growth no longer has to mean burning more carbon. Heavy industry - steel, cement, chemicals, paper - runs on constant heat and power. Today that mostly means fossil fuels. A 20-megawatt reactor that shows up on a truck is, in this telling, a way to give those industries a carbon-free baseload they can actually buy.
The skeptic's note belongs here, in fairness: no commercial PWR-20 is generating grid power yet, regulators move at their own pace, and nuclear's history is littered with confident timelines. Last Energy's bet is that standardization and private financing change the odds. The next two years - Texas criticality, UK licensing - are where the thesis meets the meter.
Return to that concrete pad outside Bryan. If the pilot reaches criticality on schedule and the UK site clears its license, Last Energy will have done something the nuclear industry has talked about for decades and rarely delivered: turned a reactor into a repeatable, shippable, financeable product. Not a moonshot. A catalog item.
That is the whole wager. Power demand is exploding, the grid is straining, and the cleanest constant source of energy we have is still treated as a once-a-generation megaproject. Last Energy is betting the fix is not a bigger reactor - it's a smaller one you can order, truck in, and switch on. The flatbed pulling onto that slab in Texas is either the start of that idea working, or a very expensive rendering. The next twelve months decide which.