Fusion power might finally work. Getting cheap is another story.

Fusion power has been the great “almost here” technology for decades. But assuming the engineering hurdles get solved—and that’s a big if—there’s another question that doesn’t get enough attention: will it actually be affordable?

A new paper in Nature Energy pours cold water on the idea that fusion will follow the same cost trajectory as solar panels or lithium-ion batteries. The researchers crunched the numbers on something called the experience rate—basically, how much costs drop every time deployment doubles. Solar modules have a 23% experience rate. Lithium-ion batteries sit at 20%. Onshore wind is around 12%.

Fusion? The study estimates somewhere between 2% and 8%. For context, traditional fission nuclear power has been stuck at 2% for decades. So fusion might beat fission, but not by much, and certainly not enough to compete with renewables on price in any reasonable timeframe.

The logic is straightforward. The researchers looked at three factors that correlate with how fast a technology gets cheaper: unit size, design complexity, and how much customization each installation needs. Fusion plants will be massive—think coal or fission plant scale. They’ll require less customization than fission because safety regulations should be simpler, but far more than solar panels. And on complexity? The experts they surveyed essentially said fusion is off the charts. One literally couldn’t fit it on the scale the researchers provided.

That’s not surprising. A fusion reactor has to contain plasma at millions of degrees, maintain magnetic confinement or laser precision, and handle neutron bombardment that degrades materials over time. It’s not a solar panel you can bolt onto a roof.

Lingxi Tang, a PhD candidate at ETH Zurich and one of the study’s authors, puts it bluntly: “If you’re talking about decarbonization of the energy system, is this really the best use of public money?” The US allocated over $1 billion to fusion in fiscal year 2024, and private-sector funding hit $2.2 billion between July 2024 and July 2025. That’s real money that could go to deploying wind, solar, storage, or grid upgrades today.

But here’s where I push back a little. The study’s approach is solid, but it’s also inherently backward-looking. You’re asking experts to estimate characteristics of a technology that doesn’t exist yet, then extrapolating cost curves. Egemen Kolemen from Princeton Plasma Physics Lab makes the point that in 2000, analysts thought solar would stay expensive too. Then China went all in, production exploded, and prices cratered. “People weren’t exactly wrong then,” he says. “They were just extrapolating what they saw into the future.”

So yeah, the 2-8% estimate is a reasonable baseline. But it assumes fusion development proceeds more or less like other large-scale energy projects. What if a breakthrough in materials science simplifies reactor design? What if modular approaches—like the smaller tokamaks some startups are pursuing—change the unit size equation? What if regulatory frameworks are designed from scratch to avoid the cost overruns that plague fission?

Those are unknowns, not certainties. But they’re also not fantasy. The fusion landscape today looks nothing like it did ten years ago. Private companies are pursuing radically different designs, from stellarators to inertial confinement to magnetized target fusion. Some of these might have fundamentally different cost structures.

Still, the study is a useful reality check. Fusion hype has been running hot, and the assumption that it’ll automatically get cheap with scale is exactly the kind of thinking that leads to bad policy and wasted investment. If fusion does work, it’s likely to be expensive for a long time. That doesn’t mean we shouldn’t pursue it—just that we should be honest about what we’re buying.

This article is from The Spark, MIT Technology Review’s weekly climate newsletter.