A Fusion Startup Just Raised $121 Million to Solve One of Fusion Power's Biggest Problems: Making Its Own Fuel
Kyoto Fusioneering has raised $121 million and is moving its U.S. headquarters to Oak Ridge, Tennessee, where it will build a device called Unity-3 in partnership with Oak Ridge National Laboratory (ORNL). The company received grants from the U.S. Department of Energy and the state of Tennessee to support the project, which it describes as the world's first facility demonstrating a breeding blanket (TechCrunch).
A breeding blanket is a component that wraps around the core of a fusion reactor and does two essential jobs at the same time. It captures the intense heat the reactor produces, so that heat can be used to generate electricity. And it makes fresh fuel for the reactor.
Here is how that works in the design Unity-3 will test. The blanket is made of liquid lithium. Inside a running fusion reactor, tiny particles called neutrons fly outward at high speed. When those neutrons hit lithium atoms, the lithium splits into helium and tritium. Tritium is a form of hydrogen that many fusion reactors use as fuel. So the blanket both absorbs heat and produces the very fuel the reactor needs to keep running. Without a working breeding blanket, a fusion power plant cannot supply its own fuel or capture useful energy. It is a prerequisite, not an upgrade.
Unity-3 will test liquid lithium alongside several other candidate materials. The results will not be kept to Kyoto Fusioneering alone. Four other fusion startups — Realta Fusion, Thea Energy, Type One Energy, and Xcimer Energy — will use the experimental data to guide their own reactor designs. According to a Fusion Industry Association survey, over half of fusion startups plan to work with outside suppliers like Kyoto Fusioneering on fuel technology rather than building it themselves.
The choice to build at ORNL puts Kyoto Fusioneering inside an existing hub of nuclear expertise. Oak Ridge has hosted tritium-related research and materials science programs for decades. Moving the U.S. headquarters there signals a long-term commitment, not just a temporary research partnership.
Kyoto Fusioneering is a spinout from Kyoto University. It designs and tests a range of technologies for fusion plants, including plasma heating, fuel systems, and heat conversion. Rather than building a reactor of its own, the company positions itself as a supplier that serves the entire fusion industry.
The company has been busy on other fronts as well. In July 2026, Kyoto Fusioneering and NGK announced a partnership on molten salt technology for fusion plants. In May 2026, the company received an order from QST to build a mockup of a tritium recovery and measurement system for ITER, the international fusion project. In April 2026, it signed a joint research agreement with the University of Tokyo and Starlight Engine. Its European subsidiary has received over 2.5 million euros in grants from the German Federal Ministry of Education and Research (BMBF) for fuel cycle equipment and lithium enrichment technology.
The UK government has also invested in this area. In January 2025, the UK Atomic Energy Authority (UKAEA) announced £9 million in funding for 12 small-scale tritium breeding and digital simulation experiments. UKAEA's Integrated Engineering division has a collaboration with Kyoto Fusioneering to develop a fusion breeder blanket concept using computer modeling. Oxford Sigma, Kyoto Fusioneering, and the UK STEP (Spherical Tokamak for Energy Production) programme announced a collaboration in December 2024.
Closer to the company's home base, Fusion Fuel Cycles (FFC), a joint venture involving Kyoto Fusioneering, is scheduled to begin operating in late 2026. At that point, integrated testing will start under controlled tritium conditions. Kyoto Fusioneering has also stated that its UNITY-1 facility will generate electricity from heat captured in liquid metal in 2026, simulating the conversion of fusion energy to usable power. General Fusion signed a memorandum of understanding with Kyoto Fusioneering in October 2023.
What this adds up to is a coordinated, multi-country effort to build the infrastructure a fusion power plant needs — before any commercial reactor has actually produced more energy than it consumes. The breeding blanket, fuel recovery, and heat extraction systems are being developed in parallel across Japan, the U.S., the U.K., and Germany, with Kyoto Fusioneering as a common thread through all of it.
The broader context here is that the fusion field has historically focused most of its attention and money on the plasma — the superheated gas where fusion reactions happen — because that is where the physics is hardest. Breeding blankets, fuel handling, and heat extraction are less exciting problems, but they are the ones that determine whether a reactor that achieves fusion can also deliver electricity to a grid. The $121 million raised by Kyoto Fusioneering, along with government grants from three countries, suggests that investors and public agencies have come to the same conclusion.
The risk, as with any technology this early, is that the reactor designs relying on this data may change in ways that make the blanket design less useful. A breeding blanket built for one type of reactor may not work well in another. Unity-3's decision to test multiple materials is, at minimum, a way to hedge against that uncertainty.
