Analysis
Amazon will fund development of a proposed 960-megawatt nuclear plant at a demolished coal-plant site south of Terre Haute, Indiana, and will be the first buyer of its power, the company and reactor maker X-energy said Monday, according to TechCrunch. The plant would use 12 of X-energy's Xe-100 small modular reactors, with American Electric Power operating the facility.
It's Amazon's second commercial agreement with X-energy and roughly three times the size of its first, a deal built on top of Amazon's lead role in X-energy's $700 million Series C-1 round in 2024. The power would feed the public grid, with Amazon as the anchor buyer -- the company says it would also serve its own data centers, though Monday's announcement stopped short of framing the deal explicitly as an AI-power play, instead pitching small modular reactors as a way to revive an industry that has struggled for decades with cost overruns and construction delays.
“nuclear capacity by 2039, a goal that depends on X-energy first successfully delivering its flagship project at Dow Chemical's Seadrift site in Texas.”
The companies are aiming for 5 gigawatts of new U.S. nuclear capacity by 2039, a goal that depends on X-energy first successfully delivering its flagship project at Dow Chemical's Seadrift site in Texas. That project hasn't broken ground: its construction permit is in the final phases of Nuclear Regulatory Commission review, and the site is tied up in a Texas lawsuit alleging "habitual non-compliance" with water and waste regulations. Mass manufacturing of SMR components is supposed to eventually lower costs industry-wide, but the benefits of that scale are likely a decade away even in optimistic timelines.
Markets read the announcement cautiously: X-energy shares, which IPO'd in April at $23 and have since fallen nearly two-thirds from their post-debut high, rose about 10% in pre-market trading on the Indiana news but gave back most of that, trading down less than 2% once the news was fully priced in. That gap between hyperscaler enthusiasm and public-market skepticism is the clearest signal in the story: Amazon is willing to commit capital and offtake years ahead of any reactor actually generating power, while investors who can trade X-energy daily are pricing in real execution risk.
Seadrift, not Indiana, is the real signal to track -- until X-energy's first flagship reactor clears its NRC permit and breaks ground in Texas, every subsequent deal, however much larger, is a bet on a timeline nobody has yet proven out.