Three nuclear startups have brought test reactors to criticality under a Department of Energy pilot program, giving the Trump administration a made-for-July-4 proof point for its push to speed advanced nuclear development.
The companies are Valar Atomics, Antares Nuclear and Deployable Energy, according to the Department of Energy and reporting by Wired. Criticality means a reactor is sustaining a nuclear chain reaction. That is a necessary step for a working reactor, but it does not mean the machine is producing useful electricity, ready for customers or licensed for commercial deployment.
Energy Secretary Chris Wright has described the effort as part of “America’s nuclear renaissance.” President Donald Trump’s May 2025 executive order set an aggressive deadline to get at least three reactors critical around the country’s July 4 250th anniversary celebrations.
Adam Stein, director of the Nuclear Energy Innovation program at the Breakthrough Institute, told Wired the prototypes are both meaningful and limited. He said the milestone is valuable for the companies and for investor confidence, while stressing that these are test reactors rather than commercial products.
What the pilot changed
US nuclear power has long been built around large light-water reactors, which use water to transfer heat and help control the nuclear reaction. Startups have pitched smaller reactors using other designs as a way to supply round-the-clock carbon-free power, including for data centers. The pitch has attracted Silicon Valley investors and tech executives, who have pressed the Trump administration to cut regulatory delays.
The administration has responded by moving parts of the process out of the usual slow lane. In February, the Department of Energy reduced several environmental and safety requirements for reactors under its authority, including pilot-program reactors, according to Wired. Similar changes are being developed at the Nuclear Regulatory Commission, which handles reactors intended for commercial sale.
Stein said shorter reviews, including for environmental impact statements that can take years, saved companies substantial time. Matt Loszak, cofounder and CEO of Aalo Atomics, told Wired that government priority changed the pace of approvals, saying signatures that once might have sat for weeks were now handled the next day. Aalo is in the pilot program but had not yet reached criticality, though the company expects to do so soon.
Labs helped get the reactors there
The startups did not reach this point alone. Several companies in the pilot program have used federally funded national laboratories. Valar Atomics reached criticality late last year at Los Alamos National Laboratory using its own fuel in a core that also included key structural components from the lab. Wired reported that Valar later reached criticality again with a second reactor at a state-funded lab site in Utah.
Antares Nuclear and Deployable Energy also reached criticality at national labs, according to Wired. That matters because a lab achievement with government support is not the same as a repeatable factory product with a fuel supply, customers and regulatory approval.
A reactor can be critical without generating electricity. Aalo’s test reactor, for example, lacks the sodium component planned for the company’s commercial design. Valar’s reactor design powered an Nvidia chip during a short demonstration, Bloomberg reported, making it the first advanced reactor in the US to provide electricity. A demo is still a demo.
The hard part comes next
Commercial reactors will still need NRC licensing, a process that has historically taken years. Wright told CNBC that the NRC is working with the Department of Energy on a faster commercialization path for companies in the pilot program.
Stein also pointed to supply chains, especially fuel, as a major obstacle for startups trying to turn lab units into products. That issue is sharper for companies that have relied on the Department of Energy for fuel access.
Brett Rampal, senior director of nuclear and power strategy at Veriten, told Wired that reaching criticality with new reactor technology in 2026 is an impressive achievement. Rampal, whose firm counts Aalo as a client, also warned that the industry can get too sentimental about a nuclear revival while downplaying the familiar math: US nuclear plants have often taken longer and cost more than planned.
This story draws on original reporting from WIRED.
