This $470mn nuclear startup is racing a presidential deadline, not the AI boom

Antares has raised $470mn in a Series C round co-led by Paradigm and Caffeinated Capital to build microreactors for US military bases, racing to meet a 30 September 2028 deadline set by Executive Order 14299. The startup's reactors produce no more than one megawatt, far smaller than typical microreactors, and it achieved criticality with its Mark-0 reactor at Idaho National Laboratory in June, the first privately developed non-light-water reactor to do so in the US in over four decades. CEO Jordan Bramble says the company has firm contracts and a named site at Joint Base San Antonio, with more military projects expected this year.

Most of the money pouring into small nuclear reactors is chasing AI data centres. This round is chasing a date. Antares has raised $470mn, co-led by Paradigm and Caffeinated Capital, with Point72 Ventures, Shine Capital and Industrious Ventures joining. The split is $370mn of equity and $100mn of debt, Bloomberg reported https://www.bloomberg.com/news/articles/2026-07-27/startup-raises-470-million-to-build-tiny-reactors-for-military . The company was founded in 2023 and has now raised more than $600mn. It wants reactors on US military installations by 2028. The deadline doing the work Executive Order 14299 directs the Department of War to have a reactor operating at a domestic military installation by 30 September 2028. That is the clock. Antares has firm contracts, a named site and a fixed date, which is an unusual position for a nuclear company of any size. The company states the reasoning plainly in its announcement https://www.businesswire.com/news/home/20260727569769/en/Antares-Raises-470M-Series-C-to-Deploy-Nuclear-Microreactors-for-Critical-Missions/ . Many US bases draw power from a commercial grid under strain from rising demand, extreme weather and, as it puts it, adversary targeting. The Air Force picked Antares under its ANPI initiative https://thenextweb.com/news/us-microreactor-companies-air-force-bases-nuclear for a proposed deployment at Joint Base San Antonio. Chief executive Jordan Bramble says more military projects will be announced this year. How small is small The scale is easy to miss and it is the most interesting number here. Antares reactors produce no more than one megawatt. A conventional nuclear plant makes about one gigawatt. Even among microreactors, which usually top out around 20 megawatts, Antares sits at the very small end. A factory builds these units and ships them whole. They run for years without refuelling, and the company targets six years or more of deployed operation. That is a different product from the reactors being built for the AI boom. Valar Atomics raised $450mn https://thenextweb.com/news/valar-atomics-nuclear-ai-data-centre-funding to power data centres. X-energy raised $1.02bn https://thenextweb.com/news/x-energy-ipo-billion-nuclear-ai-data-centres in a Nasdaq listing. In the UK, Nuclear Turbines raised £15mn https://thenextweb.com/news/nuclear-turbines-15m-raise-compact-reactors on Monday to make industrial power cheaper. Antares is selling resilience to soldiers, not cheap electricity to hyperscalers. The milestone, and how to read it In June, Antares took its Mark-0 reactor critical at Idaho National Laboratory, meaning it sustained a fission reaction. The company calls this the first privately developed non-light-water reactor to reach criticality in the United States in more than four decades. Paradigm’s Alana Palmedo went further, calling it the first private advanced reactor criticality in decades. Bloomberg frames it more narrowly. It reports Antares as the first participant in the Energy Department’s reactor pilot programme to produce a self-sustaining fission reaction. That is a real first, and a smaller one than the company’s wording implies. The test ran on TRISO fuel. It validated reactor physics, reactivity control and instrumentation in a full-scale core, and it produced only a small amount of energy. “On June 4th, we won the race to criticality, and now we’ve shifted to the race to commercialisation,” Bramble said. What has not happened yet No microreactor is in service anywhere in the United States. Bloomberg notes the technology remains unproven at commercial scale. Antares does not expect a system that generates electricity until next year, when the Mark-1 is due. Criticality and useful power are separate achievements, and the company has one of them. Bramble frames the raise as matching public money rather than replacing it, describing the $470mn as investing “one-to-one with the taxpayer” to reach commercial scale. The bottleneck nobody mentions Fuel is the constraint that has strangled advanced nuclear for years, and Antares moved on it in May. It signed a long-term enrichment agreement with Urenco for High-Assay Low-Enriched Uranium, which Urenco called the world’s first multi-year contract for HALEU supply. Advanced reactors need HALEU. Western supply of it has been thin, and Russia was for years the most reliable commercial source. Locking in enrichment is arguably a harder problem than locking in capital, and it attracts a fraction of the attention. Why this one is worth watching Britain and Europe are running the same experiment on a different footing. A Polish billionaire’s £35bn plan for 14 mini reactors https://thenextweb.com/news/solowow-sge-35bn-uk-small-modular-reactors-bwrx-300 targets first power in 2034, six years after Antares plans to be switched on at a US airbase. The gap is not technical so much as procedural. American defence procurement has produced a hard deadline, a named customer and a signed contract, which is a faster mechanism than a competitive energy market. Whether the reactor works on time is a separate question. Antares has made a fission reaction happen in a laboratory in Idaho and has not yet made a watt of useful electricity anywhere. Get the TNW newsletter Get the most important tech news in your inbox each week.