Four advanced reactor developers met an unusually aggressive federal deadline by bringing test reactors to criticality before or on July 4, 2026, a milestone Antares, Valar Atomics, Deployable Energy, and Aalo Atomics each cleared it in turn. None of the four reactors produces electricity; all were zero-power demonstrations that validated a fuel-and-core design and DOE's authorization process, not commercial heat production or sustained output. What that coverage did not have yet is what has happened to these companies' balance sheets since: three of the four are now raising, or have just closed, financing rounds an order of magnitude larger than anything that got them to criticality. That gap between the technical milestone and the capital now required is the real story for anyone underwriting the next phase of advanced nuclear.
DOE Built a Faster Test Path, Not a Faster Commercial Path
The Reactor Pilot Program moved these demonstrations through DOE's own authorization process rather than the Nuclear Regulatory Commission licensing pathway used for commercial deployment, at DOE and affiliated sites, mainly Idaho National Laboratory, where existing nuclear infrastructure removed barriers a new commercial location would still face. Developers covered their own project costs, but the demonstrations also drew on federal authorization, laboratory infrastructure, and in some cases federally allocated fuel. That combination shows how fast a test reactor can move when fuel, land, and federal oversight are coordinated around one deadline. It does not establish that a commercial plant can move at the same speed once developers face full NRC licensing, site development, transmission, and construction from scratch.
The Capital Already Flowing In Shows the Scale of What's Next
The companies that hit criticality are not operating on seed money. Aalo Atomics raised a $100 million Series B in August 2025, bringing its total to $136 million at the time, and the company's own website now puts total funding above $300 million. Antares took a different path to a similar place: a $96 million Series B last December, followed by a $470 million Series C on July 27, split between $370 million in equity and $100 million in debt, pushed its total funding to roughly $604 million. Valar Atomics may be about to leapfrog both of them. The company is reportedly in talks to raise $1 billion at a $6 billion valuation, with Sequoia Capital in discussions to lead, on the strength of a reactor that has already sent a trickle of power to an Nvidia AI chip through a new partnership.
Those are substantial raises for advanced nuclear startups, and they fund reactor engineering, testing, licensing preparation, fuel procurement, and early manufacturing. They are still not the same as financing a commercial nuclear project. The next stage requires combining that corporate capital with government procurement, utility investment, project finance, customer commitments, or regulated cost recovery, an assembly problem rather than simply a bigger venture round.
Commercial Nuclear Requires a Different Scale of Capital
First-of-a-kind commercial nuclear projects routinely require capital measured in billions of dollars, though direct comparisons across reactor designs are difficult. Ontario Power Generation estimates the first 300-megawatt BWRX-300 reactor at its Darlington site will cost roughly $4.4 billion (C$6.1 billion), with another $1.2 billion (C$1.6 billion) budgeted for shared infrastructure, and the full four-reactor program estimated at $15.1 billion (C$20.9 billion) including interest, escalation, and contingency. Darlington is a far larger reactor design than the microreactors most Reactor Pilot Program participants are building, so it is not a direct comparison. It still illustrates the underlying problem. Moving from a small test reactor to commercial power means a licensed site, a secured fuel supply, cooling and electrical systems, a construction workforce, and an operating organization built to run for decades, not just a bigger reactor core. Each of those pieces adds its own cost and schedule risk on top of the reactor itself.
NuScale's canceled Carbon Free Power Project is the clearer warning for smaller designs: its estimated cost rose from $5.3 billion to $9.3 billion before NuScale and its utility customers terminated the project after failing to secure enough subscriptions. Technical progress does not automatically produce a financeable electricity price.
Investors Are Underwriting Two Different Capabilities
The criticality milestone shows a company can develop a working test configuration, manufacture or procure reactor components, and move through DOE's authorization process under controlled conditions. Commercial deployment asks something broader: whether the same company can finalize a commercial design, win regulatory approval, lock in fuel supply, build a repeatable manufacturing line, control construction costs, secure insurance and financing, sign creditworthy customers, and operate the plant for decades. Those are not simply larger versions of the same task; they require different personnel, partners, and sources of capital. An investor funding a test reactor is underwriting technology development. An investor funding a commercial plant is underwriting a regulated infrastructure project with construction, market, and counterparty risk attached, the same execution gap already reshaping how companies plan capital-intensive infrastructure bets across other sectors this year.
Data Center Demand Doesn't Close the Financing Gap
Advanced nuclear developers increasingly point to data centers and AI infrastructure as anchor customers, and the logic holds: large digital facilities need dependable, round-the-clock power and are running out of grid capacity to get it. But demand for electricity is not the same as a bankable reactor contract. Aalo's own CEO has pitched fleets of small reactors as dedicated, zero-carbon power for AI infrastructure, and a data center operator weighing that pitch still has to price in whether the reactor licenses on schedule, whether construction costs are fixed or exposed to escalation, who covers replacement power during an outage or delay, and whether the final delivered price stays competitive once all of that is accounted for. Some developers will pursue conventional power purchase agreements; others may lean on government procurement, build-own-operate structures, or direct onsite supply deals. Whichever model wins out, buyers will need more than a successful criticality test before they sign.
Commercial Scale Is the Real Test
DOE and the four developers proved that small test reactors can move through authorization and reach criticality on an unusually compressed schedule when a program is organized around a hard federal deadline. That is a real accomplishment. It did not prove that commercial reactors can be licensed and built at the same pace, establish a commercial cost curve, or show that customers will accept the resulting electricity price. The developers that separate themselves from here will not necessarily be the ones that reached criticality first. They will be the ones that can turn that milestone into a licensed commercial design, a committed customer, and a financing structure strong enough to survive the years between a successful test and the first delivered megawatt.
