U.S. small modular reactors from vendors including GE Vernova, NuScale, Bloom Energy, and Cameco are positioned for their strongest year in 2026, driven by demand from AI data centers.

In 2023, NuScale's flagship small-reactor project collapsed under soaring costs, and the central question around nuclear power has since shifted to whether small modular reactors are finally becoming real or whether 2026 is just another round of expensive optimism. For utilities, taxpayers, and households facing rising electricity demand, the answer carries major implications. If SMRs work, they could add reliable, low-pollution power to the grid. If they don't, they could siphon time and money from cheaper options already scaling up.

2026 may be the most encouraging year yet for small modular reactors, marked by a burst of activity. Several commercial nuclear projects are starting construction after about a decade without new groundbreakings, and the Trump administration has reserved $900 million for SMRs, with $94 million already awarded. The renewed interest is notable partly because NuScale's Utah project failed so publicly: projected costs rose from $3 billion in 2015 to $4.2 billion in 2018, then $6.1 billion in 2020 and $9.3 billion in 2023, before customers backed away.

SMRs are intended to produce about 70 to 350 megawatts—compared with the 600-to-1,500-megawatt output typical of conventional reactors—with major equipment built in factories and assembled later on-site. As one analyst framed it, the goal is to "turn nuclear power from a construction industry back into a manufacturing industry." Yet that business case assumes manufacturing scale the sector has not achieved. SMR power currently costs roughly $214 per megawatt-hour before any factory-scale learning effects, compared with about $40 to $98 for utility-scale solar and $37 to $99 for onshore wind.

In the U.S., nuclear still provides about one-fifth of electricity even though much of the reactor fleet was built before 1990. More nuclear capacity could cut reliance on coal, oil, and gas and provide steadier support for the grid as electricity demand rises. But nuclear power carries the familiar downsides of very high upfront costs, long construction periods, and radioactive waste. Some SMR concepts face fuel-supply and regulatory limits, making deployment more complicated than optimistic headlines suggest.

With more than 100 SMR designs now under development worldwide, it may be harder for any single design to be repeated often enough to bring costs down. Even nuclear supporters acknowledge the competitive pressure from solar and battery storage, which continue to get cheaper and easier to deploy. Recent progress includes partnerships to advance tiny reactor development, emerging technical solutions for molten-salt reactors, and research tackling corrosion and leak risks in next-generation designs.