By TechCrunch's count, more than a dozen private companies have raised in the ballpark of $10 billion to build commercial fusion power plants. The hard part is no longer proving the physics works; it is building the magnets, fuel cycle, and supply chain a working plant requires.
Private fusion has crossed a capital threshold that would have looked absurd five years ago. In a Friday roundup, TechCrunch's Tim De Chant counted every fusion company that has raised more than $100 million and found roughly a dozen names, with cumulative private capital in the same neighborhood as the gross domestic product of a small country. The most prominent example is Commonwealth Fusion Systems, which has raised close to $3 billion, per De Chant's reporting, and is building the Sparc demonstration plant in Massachusetts.
The capital is the headline. The story is what it actually means. For decades, "fusion" carried the punchline of being twenty years away and always would be. That joke is not yet dead, but the money is now betting that it should be. The shift is not about new physics. It is about new engineering. In December 2022, the U.S. Department of Energy's National Ignition Facility produced more energy out of a controlled fusion reaction than its lasers put into the fuel pellet, the first time a laboratory cleared scientific breakeven. De Chant's coverage cites the milestone as the marker of the moment, but the 2.05 megajoules of laser energy in and 3.15 megajoules of fusion energy out belong to a national-lab target chamber, not a power plant.
What that result actually settled is the question of whether fusion can work in principle. It did not settle whether anyone can build a fusion power plant. That is a much harder question, and it is the one the current capital is buying. Putting a fusion plant on the grid requires a stack of capabilities that no single company can fund from a venture round: high-temperature superconducting magnets strong enough to confine plasma for minutes at a time, a tritium fuel cycle that produces and recaptures its own fuel, blanket technology that converts fusion neutrons into heat, and grid interconnection engineering that lets a plant behave like a baseload generator rather than a research project.
The companies on TechCrunch's $100 million list are chasing those shared problems along different technical paths. Tokamak Energy, Helion, and Commonwealth Fusion Systems are pursuing magnetic confinement, the family of reactor designs that uses powerful magnetic fields to hold a ring of plasma in place, in different shapes. TAE and Realta are working on field-reversed configurations and magnetized target fusion, which mix magnetic fields and physical pistons to compress plasma. Xcimer, Pacific Fusion, and Avalanche are betting on inertial confinement and pulsed-magnetic approaches, which use lasers or fast-rising magnetic fields to ignite short, controlled bursts. Kyoto Fusioneering, Marvel Fusion, Shine, and Type One are pursuing adjacent pieces of the problem: fuel cycle engineering, hybrid inertial confinement, and medical isotope production as nearer-term revenue. The diversity is not chaos. It is hedge. No one yet knows which approach, if any, will produce electricity at a price the grid will accept.
Three enabling advances make the current bet possible. The first is high-temperature superconducting tape, which has finally become good and cheap enough to build the kind of magnets a compact tokamak needs. The second is AI-accelerated plasma simulation, which lets engineers run thousands of design variations in silico before committing to a single magnet geometry. The third is the late-2022 NIF result, which gave capital a single, dated event to point at as evidence that the underlying physics is real.
The honest caveat is that scientific breakeven is not commercial breakeven. A laser pulse that delivers 3.15 megajoules inside a national-lab target chamber is not a power plant. Capital intensity is high. Timelines are long, and several of the companies on this list have raised large sums precisely because the engineering is hard. The NIF result is a tailwind, not proof of fusion power economics. TechCrunch's funding totals reflect a curated aggregator's tally rather than a primary funding database, and individual figures will need to be checked against company press releases before they are treated as settled.
What to watch next is execution, not demos. The signal that the industry has genuinely moved from the demonstration phase to the build phase will not be another scientific record. It will be a magnet factory coming online, a tritium supply agreement signed, or a grid interconnection deal with a real utility. Until then, the $100 million club is a bet. The bet is that shared engineering, not competitive physics, is the next constraint to fall.