Who’s Really Ahead in the Race To Fusion Energy? Background for Journalists

The merger announced on December 18, 2025, between the Trump Media and Technology Group and the fusion company TAE was widely reported, but the reactions have been widely varying. Some have described it as a major financial breakthrough for fusion, others as a pump-and-dump scheme to fleece small investors. As part of the merger announcement, TAE announced plans to begin in 2026 construction on a fusion power plant that will actually produce electricity for the grid by 2031. TAE has now become the fourth fusion company to claim that they have begun construction on an operating power plant, along with CMS, Helion and Helical Fusion in Japan.

But are any of these claims at all credible from a technological standpoint? Sure, these companies have billions of dollars in investments. But does that make them leaders in the race to actually achieve fusion energy? Who really is ahead in the race to develop fusion energy? This release provides a brief guide for non-technical journalists.

How do we measure who’s ahead?

First, what are good measurements to tell who is ahead? Most media use money invested as a measure of who the leading fusion companies are. But that is a really poor measure of who is close to the goal of commercial fusion energy or even of who is getting close. Lots of money can be wasted on technological directions that just don’t work. The ITER project is a good example, as it is by far the fusion project that has had the most money invested in it– about $25 billion. After decades of preparation, the construction of the pyramid-scale device was initiated in 2006 and was planned to last for 10 years. Fusion operation was planned for 2023, 17 years in the future. This was not to be commercial energy generation, just more energy out of the device than is put into it, a big step called “net energy”.

But by 2024, the completion date was put off until 2034, still ten years in the future and fusion operation to 2039, 15 years in the future. Thus, the expenditure of tens of billions of dollars has not brought the start of this huge experiment (and it is an experiment) any closer.

The fact that public, not private, dollars were involved does not at all affect the picture. Helion, for example, in 2014 announced that it would be delivering a commercial fusion generator in 2019. Despite 1 billion in funding, Helion in 2024 announced that it was building a commercial fusion generator to begin operation in 2028. A decade and a billion dollars had not brought its estimates closer to reality.

These large projects are not the only ones predicting faster progress than materializes. Smaller companies, including LPPFusion, have also been overly optimistic. The point is that billions of dollars in expenditures are not a predictor of eventual success.

There are two good measures of actual fusion achievement and of how close a project is to success. The best one is “wall-plug efficiency”. The other one is n𝛕T measure, and we will describe this below.

Wall-Plug Efficiency

The other one is n𝛕T measure, and we will describe this below.

Deuterium

The most widely used experimental fusion fuel is deuterium by itself—it’s cheap, non-radioactive and easily available. For deuterium wall-plug efficiency, the clear leader is not the billion-dollar companies, but LPPFusion. In 2025, LPPFusion broke its own records to achieve a wall-plug efficiency of 4 J of fusion out for every million J (MJ) in. (The experiment put 60 kJ into the device and got a bit more than ¼ J out.) That is far from net energy, but no one expects to reach net energy with deuterium by itself.

The next closest company is another small firm, MIFTI. They achieved a bit more than half as much fusion out as LPPFusion, but with almost ten times as much energy in, so their wall plug efficiency is only ¼ J out per MJ in, a factor of about 15 less than LPPFusion. (They put 600 kJ into the machine and got 0.15 J out). MIFTI uses a device called a z-pinch, a close cousin of LPPFusion’s dense plasma focus device.

For comparison, the best result with deuterium alone achieved by government-funded labs is only a small factor larger than the records achieved by LPPFusion. JET achieved a fusion yield of about 50k J with a total energy input of about 10 GJ, for a wall plug efficiency of 5 J per MJ in, just 25% more than LPPFusion.

Hydrogen Boron – pB11 fuel

For the other main fuel that private companies are experimenting with, hydrogen-boron, the leaders are different. HB11, in Australia, has achieved 0.18 J per MJ with a laser beamed directly onto a target. TAE used a stellarator in Japan, not its own different device, to achieve 0.012 J per MJ with a hydrogen-boron plasma. For comparison, a university lab in China, Xi’An Jiatong University, achieved a much higher wall-plug efficiency of 3.5 J per MJ with a somewhat different laser concept called ”pitcher-catcher.” In this approach, the laser is used to generate a high-intensity proton beam. This beam then hits a boron target.

Deuterium – Tritium or DT

For comparison, the big national laboratories working with a third fuel – deuterium-tritium or DT– have achieved much higher wall-plug efficiencies. The recently-shut JET tokamak in the UK achieved a wall-plug efficiency of 0.7% with DT fuel and, of course, the huge laser NIF in the US has done even better, achieving a wall-plug efficiency of 2%.

The n𝜏T measure

The wall-plug efficiency measure can’t be applied to all companies, since some have experimented only with hydrogen, which produces negligible fusion, or have reached such low temperatures that deuterium does not fuse either. For all companies, there is another, somewhat rougher measure, called n𝜏T. This is the product of the density of the plasma, the confinement time and the temperature. Each of these factors are required to get to net energy. But it’s a rough comparison, because the fusion output does not depend linearly on temperature.

By this measure, LPPFusion clearly is in the lead among private companies. LPPFusion’s experiments, verified in peer-reviewed publications, have achieved a product of 3.4 x 1020 keV-s/m3. MIFTI is in second place with 2.6 x 1019 keV-s/m3, a factor of 13 less, Tokamak Energy is third with 6 x 1018 keV-s/m3and TAE is fourth with 6×1016 keV-s/m3. Again, by this measure, the two leading companies have raised only about $12 million apiece but have achieved results that are far better than companies like TAE with hundreds of times more money invested.

Why money alone does not mean fusion

Why is there such a complete disconnect between results and money invested? Quite simply, there are paths to fusion that are a lot cheaper and other paths that are much more expensive. Paths that use low-density plasmas and external magnetic fields, such as tokamak or the field-reversed configurations that both TAE and Helion use, necessitate large machines that are consequently costly. Approaches that use high-density plasmas and only an internal magnetic field produced in the plasma by current passing through it, including the dense plasma focus used by LPPFusion, the z-pinch used by MIFTI and the laser approaches used by HB11, allow the use of small machines that are far cheaper. This is particularly true for the DPF and z-pinch, since in these approaches, the energy fed into the device is always in the form of electricity. With lasers, the low-efficiency conversion of electricity into light leads to somewhat greater expense in the lasers themselves, although not as great as with low-density machines like the tokamak or the stellarator.

So, the scientific story is that the leading companies in fusion are LPPFusion, HB11 and MIFTI, not the financial leaders like TAE and Helion. The company that develops fusion first will dominate the market, not the ones who have spent the most money.

However, most media also get an important point on the financial side wrong. Many have reported that the merged TMTG-TAE is the only “pure fusion” play for small investors. That’s wrong in two ways. TMTG is far from a pure fusion play since the firm’s value also depends on the value of Bitcoin and the political value of the Trump connection. In addition, small investors can invest in LPPFusion through the Wefunder crowdfunding platform. That is a pure fusion play, as LPPFusion has no other business.

Addendum:

Can TAE Start Building a Working Fusion Generating Plant in 2026?

As part of the merger announcement, TAE announced plans to begin in 2026 construction on a fusion power plant that will actually produce electricity for the grid by 2031. Is this possible? For context, TAE has now become the fourth fusion company to claim that they have begun construction on an operating power plant—so have CMS, Helion and Helical Fusion in Japan.

In our view, none of these claims are at all credible. The problem is not the near-term goal of getting electricity by 2031. Lots of fusion companies, including ours, are aiming for similar goals. The problem is that none of these four companies have completed the scientific research needed to design a fusion generator. So, none of them can start construction of something they have not and can’t design right now.

How do we know they have not completed scientific research? The universally-acknowledged goal of the research phase of fusion energy is to achieve net energy in the laboratory—more energy out of a device than is put into it. That shows that electricity production is possible in a commercial generator designed on the basis of the net energy lab experiments. None of the four companies has reached net energy, none are close to reaching it and, in fact, none are anywhere near as close as LPPFusion is.

Let’s take TAE. We can’t compare how much fusion they’ve produced with energy input because they are using pure hydrogen, not any fusion fuels, so they are not getting any fusion at all. Also, they don’t release how much energy they put into their machine. (From their papers, the input energy must be more than 10 times as much as the input into LPPFusion’s FF-2B device) However, comparisons can still be made since we all agree that net energy requires a combination of high temperature, high density and adequate confinement time. TAE has released those figures.

Compared with our best results, TAE’s are comparable for the product of density and confinement time, called n𝜏. They have 60 billion seconds x particle/cm3 and we have 90 billion seconds x particle/cm3. But they achieved that with a temperature of 1 keV (11 million K) and we got 250 KeV (2.8 billion K), or 250 times better.

To get to net energy, we both have to increase n𝜏 to around 2 quadrillion seconds x particle/cm3, so for TAE that’s a factor of 3,000 on top of the 250 times increase in temperature. That’s not even close. To put it another way, TAE’s Norm machine, their latest, uses a current of 300kA to confine the plasma. For a net energy machine, they would need about 3,000 times more or 1GA–a billion amps.

It’s not possible to expect extrapolation over so many orders of magnitude to be accurate. Far more experiments and improvements are needed before any confident prediction of breakeven.

And that’s what TAE had planned. They were going to build another bigger experimental device called Copernicus before they designed and built a power generator, which they called DaVinci in the plans. But with the merger announcement, all of a sudden Copernicus has disappeared and they are starting to build DaVinci right away.

On top of that, they are substantially altering the design of the machine—the one they call Norm is the first and only one of the new design that they’ve built.

So if, as is now planned, they are going to start building a power generator, the chance that they will make major mistakes in design is practically 100%. Instead of speeding the path to fusion, they’ll slow their progress by wasting money and time on engineering dead ends. It’s sort of as if some company in 1900 had said, ”Never mind about the control of heavier-than-air craft. We know the science and we are building a jetliner that will take 200 people across the Atlantic.”

To be fair, this is not exactly the first time that fusion management (and, more rarely, fusion scientists) have said that ”the science is done”. That was said 50 years ago, too, when the decision was made to focus on the tokamak and DT fuel. The motivation was the same—to reassure potential founders. But it was not true then and is not true now. We’ll know the science is done when some group demonstrates net energy. Then the design of a power generator can start.

Scientific references:

The different approaches to fusion energy: https://pubs.aip.org/aip/pop/article/30/12/120602/2928807/What-are-the-fastest-routes-to-fusion-energy

LPPFusion’s results: https://link.springer.com/article/10.1007/s10894-023-00345-z

MIFTI results: https://doi.org/10.1063/1.5085673