Blue Energy raises $380M for modular nuclear plants aimed at data centers

Blue Energy has raised $380M to speed up development of prefabricated nuclear plants for large energy consumers, including AI data centers. The startup is betting not on a new type of reactor but on a different way of building: more factory assembly, less complex work on site.

What the money is for

The round includes equity and debt financing. It was led by VXI Capital, with participation from Engine Ventures, At One Ventures and Tamarack Global. According to the company, the funds will go toward buying long-lead equipment, developing projects and growing the team.

Blue Energy plans to develop its first project in Texas: a site with capacity of up to 1.5 gigawatts, designed for large electricity consumers, including operators of AI data centers that already can't get enough free capacity on the grid.

Interest in such schemes has grown amid the AI boom. New data centers need enormous volumes of stable power, and grid connections in the US often stretch to years. So investors are looking not only at the reactor itself but also at how quickly capacity can come online. The timing suits Blue Energy: it can sell not just low-carbon generation but a way to get megawatts to a customer faster than in the classic nuclear construction scenario.

How the scheme works

Blue Energy is trying to solve the part of a nuclear project that has scared off investors for years: construction. Instead of assembling a plant almost entirely on site, the company wants to move a significant share of the work to existing manufacturing sites and shipyards. There it is easier to standardize processes, keep to a schedule and reduce dependence on scarce specialists at a particular construction site. The large modules can then be delivered to the site by water and installed on location.

The company stresses that it is not inventing a new reactor from scratch. Blue Energy's approach is compatible with already known reactor technologies, and the main novelty lies in the sequence of work, the contract model and the level of factory readiness of each delivery. For the industry, this is a fundamental shift from one-off projects to repeatable assembly.

The company's materials cite a target of 48 months to deploy a plant, noticeably faster than classic nuclear builds, which often run far beyond their original schedules and budgets.

"The problem with nuclear power is not the technology but the cost and predictability of construction," is how CEO

Jake Jurevich describes the logic of the project.

What the round changes

For Blue Energy this is not just a venture check for research. The round is meant to push the project toward actual construction: the company speaks of early work at the Texas site in 2026 and is aiming for a final investment decision in 2027.

There is also a separate signal for the market, a regulatory one. The NRC has already approved Blue Energy's approach to phased construction, in which part of the non-nuclear infrastructure and the turbine loop can be brought into service earlier, and the facility then switched over to nuclear generation. This matters for several reasons:

• the company is trying to make a nuclear project legible to ordinary project finance, not only to government support;
• off-site assembly should reduce the risk of schedule and budget overruns;
• capacity of up to 1.5 GW makes the project relevant specifically for hyperscale data centers;
• demand from AI infrastructure creates a rare scenario for the nuclear industry, in which a new source of baseload power is needed quickly, not "someday".

What this means

The AI infrastructure market is increasingly constrained not by chips but by electricity. Against that backdrop, Blue Energy is selling investors and customers less a "new nuclear technology" than a promise of predictable construction and clear economics. If the company really does get the Texas project to the site and holds to its schedule, it will strengthen the trend toward pairing data centers with new nuclear generation.