For decades, fusion energy was the punchline. Always thirty years away. Always promising. Never delivering. But something shifted around 2022 when the Department of Energy announced a breakthrough: scientists had finally achieved scientific breakeven, producing more energy from a controlled fusion reaction than they put in. Suddenly, the joke wasn’t funny anymore. It was a signal.
What followed wasn’t a slow burn. It was a rush. Venture capitalists flooded the space. Startups with wildly different approaches to the same problem proliferated like rabbits. And the money? The money got absurd.
Commonwealth Fusion Systems alone has vacuumed up roughly a third of all private capital invested in fusion to date. That’s $3 billion. Google signed a power purchase agreement for half the output of a commercial plant that doesn’t exist yet. Microsoft committed to buying electricity from Helion’s 2028 prototype. Even Trump Media got in on the action, merging with TAE Technologies in a bizarre all-stock deal that valued the combined company at $6 billion.
The fusion industry has become a strange beast: part serious science, part venture capital theater, and part genuine bet on the future of energy.
Why Now?
Three things converged to make fusion suddenly viable. Better computer chips meant better simulations of what happens inside a reactor. More sophisticated AI helped optimize reactor designs and control systems. And high-temperature superconducting magnets became powerful enough and stable enough to actually confine plasma without consuming more energy than they generated.
The 2022 breakthrough didn’t solve fusion. It proved the underlying physics wasn’t broken. That distinction matters. It meant startups could stop asking “is this even possible?” and start asking “how do we make it commercial?”
The problem is that scientific breakeven and commercial breakeven are not the same thing. Scientific breakeven means you got more energy out than you shot into the fuel pellet. Commercial breakeven means the entire facility produces more electricity than it consumes. That’s a different beast entirely. It’s the difference between proving something works in a lab and shipping it to your customers.
The Approaches Are Everywhere
If there’s one thing the fusion industry has in abundance, it’s philosophical disagreement. Commonwealth Fusion Systems is building a tokamak, a doughnut-shaped reactor that uses magnetic fields to contain plasma. Helion and TAE Technologies are using field-reversed configurations, where plasma is shaped like a cigar instead. Tokamak Energy is squeezing the doughnut into something that looks like a steampunk Fabergé egg. Proxima Fusion is betting on stellarators, which twist and bulge to accommodate plasma’s quirks.
Then there’s Pacific Fusion, which emerged from stealth with a $900 million Series A to pursue inertial confinement using 156 electromagnetic pulse generators that have to fire in perfect synchronization. Or Zap Energy, which uses an electric current to generate a magnetic field that compresses plasma by about a millimeter. Or First Light Fusion, which abandoned the power plant dream entirely to sell its projectile-based compression technology to other companies.
The sheer variety of approaches is, in one sense, healthy. No one actually knows which design will win. But it also means capital is spread thin, and most of these companies won’t survive. That’s not a knock on them. It’s just how venture capital works.
The Reality Check
For all the hype, fusion startups are hitting real problems. General Fusion, a company founded in 2002 and backed by Jeff Bezos, hit a cash crunch in spring 2025. The company laid off 25% of its staff and was forced to accept a $22 million “pay-to-play” round that one investor called “the least amount of capital possible” to keep the lights on. By November, the company had scraped together another $51.1 million in SAFE notes from nearly 70 investors.
First Light Fusion, which has raised $108 million, announced it would not build its own power plant. Instead, it’s pivoting to selling its core technology and building “pulsed power capability” for defense and scientific applications. In other words, it was abandoning its original mission because the original mission wasn’t going to generate near-term revenue.
These aren’t failures. They’re recalibrations. But they’re also reminders that fusion is still hard, expensive, and uncertain.
The Timelines Are Aggressive
Commonwealth Fusion Systems says its Sparc reactor will be operational in late 2026 or early 2027. Helion claims it will produce electricity by 2028. Xcimer, founded in 2022, is already building a 10-megajoule laser system and has raised $100 million. Marvel Fusion expects to have a demonstration facility operational by 2027.
These timelines should be taken with salt. The history of fusion is littered with promises pushed back by years. But the fact that serious companies with serious backing are willing to put their money behind these projections suggests something has genuinely shifted.
The Venture Capital Angle
What’s driving this isn’t just scientific progress. It’s also that fusion has become a venture capital play. A trillion-dollar market if it works. A cautionary tale if it doesn’t. That’s the kind of risk-reward profile that keeps VCs awake at night in a good way.
You’ve got some of the biggest names in tech and business behind these companies. Bill Gates through Breakthrough Energy Ventures. Sam Altman and Reid Hoffman backing Helion. Jeff Bezos in General Fusion. Peter Thiel’s Mithril Capital. BlackRock.
The money is real. The commitment is real. Whether the physics scales to commercial reality remains to be seen.
The Unsexy Play
Not everyone is betting on fusion reactors themselves. Kyoto Fusioneering is building the balance of plant, the unglamorous but essential equipment that sits outside the reactor: heat extraction systems, plasma heating tools, integration expertise. The company has raised $191 million on the bet that if even one fusion startup succeeds, someone will need to supply the supporting technology.
It’s a smart hedge. Even if 99 fusion startups fail, one might succeed. And that one will need equipment.
What’s Actually at Stake
Fusion energy isn’t just about the physics. It’s about whether we can build a scalable, economically viable alternative to fossil fuels and conventional nuclear power. Solar and wind are getting cheaper and more efficient, but they’re variable. Battery storage is improving, but it has limits. Fusion, if it works, could provide baseload power without the waste profile of conventional nuclear or the carbon emissions of fossil fuels.
The companies pushing fusion understand this. They’re not just trying to make money, though that’s certainly part of it. They’re trying to solve a problem that gets harder with every year we delay.
Whether they actually pull it off is still an open question. The timeline pressure is real. The capital is flowing. The physics is promising. But the gap between promising physics and a grid-connected power plant delivering 400 megawatts of electricity remains vast.
We’ll find out in the next few years whether these startups have finally cracked the code or if fusion remains what it’s always been: just far enough away to keep hoping, but never quite close enough to stop waiting.
