Fusion power just got a $4 billion vote of confidence, and a date to prove itself
The old joke is that fusion is always 30 years away. In 2026 the private leaders are trying to make it a dated bet instead: Commonwealth Fusion just raised its total to $4 billion, its reactor is most of the way built, and it is aiming for net energy gain in 2027. Here is what is real, and the caveat that separates a lab result from a power plant.
Fusion, the process that powers the sun, has spent decades as the energy source that is "always 30 years away." In 2026 the private frontrunners are trying to put a date on it. Commonwealth Fusion Systems raised another $1 billion in July, bringing its total to about $4 billion, while its SPARC reactor is roughly 80% assembled and aiming for "first plasma" later in 2026 and net energy gain in 2027. A rival, Helion, is chasing a different design and a 2028 deal to sell power to Microsoft. The crucial caveat: "net energy gain" in the plasma is not the same as delivering electricity to your home, and every headline date here is a target, not a result. But the money, the magnets and the machines are real.
For as long as most people have been alive, fusion power has been the punchline to a joke: the miracle energy source that is perpetually 30 years away. Clean, near-limitless, no long-lived radioactive waste, and forever just out of reach. In 2026 something changed in the tone of the conversation. The leading private companies are no longer promising fusion "someday." They are naming years. Whether they hit those years is another question, but the shift from open-ended dream to dated bet is itself the news. Here is what is actually happening, and the one distinction that keeps it honest.
What just happened
The clearest signal is money. On 30 July 2026, Commonwealth Fusion Systems (CFS), the best-funded private fusion company and a spin-out of MIT, announced it had raised another $1 billion, bringing its total capital to about $4 billion. That single company has now raised a large share of all the money the entire fusion industry has ever attracted. Investors do not put that kind of capital into a science project; they put it into something they judge has a credible path to working.
And there is a machine to point at. CFS's demonstration reactor, SPARC, is being built in Devens, Massachusetts, and is now roughly 80% assembled: the two halves of its massive vacuum vessel are installed, and its set of powerful magnets is being craned into place. This is not a rendering or a roadmap. It is steel and superconductor, most of the way to switching on.
The magnets are the real breakthrough
If you want the single technical reason fusion suddenly looks nearer, it is the magnets. A tokamak like SPARC works by holding a cloud of plasma hotter than the sun's core inside a magnetic "bottle," because no physical material could touch it and survive. The stronger the magnetic field, the smaller and cheaper the reactor can be for the same result.
CFS's bet, proven in a landmark 2021 test, was on high-temperature superconducting (HTS) magnets that reach about 20 tesla, far stronger than the magnets earlier designs relied on. That leap is what lets SPARC aim to do, in a machine you could fit in a warehouse, what the giant international ITER project is trying to do at many times the size and cost. The magnets are why "smaller, faster, privately funded fusion" went from fantasy to fundable.
The dated bet: first plasma in 2026, net gain in 2027
Here is where the specific, falsifiable claims come in, and where you should read carefully. CFS is targeting "first plasma" (the first time SPARC creates and holds a fusion plasma at all) later in 2026, and, more importantly, net energy gain in 2027.
"Net energy gain," often written Q greater than 1, means the fusion reaction releases more energy than was pumped into the plasma to sustain it, the milestone the field has chased for 70 years. A US national lab first crossed that line in 2022 using lasers; doing it in a compact, commercially oriented, magnet-based machine like SPARC would be a first of its own kind.
Now the caveat that separates honesty from hype, and it is the one almost every excited headline drops. Net energy gain in the plasma is not the same as net electricity to the grid. When a US national lab announced a fusion "breakeven" in 2022, it produced more energy than the laser delivered to the fuel pellet, but far less than the whole facility drew from the wall. A power plant has to win the entire energy balance, running magnets, cooling, and conversion, and still have plenty left to sell. SPARC hitting Q greater than 1 in 2027, if it happens, would be a historic scientific result and a validation of the approach. It would still not be a power station. The first plant CFS plans, called ARC, is targeted for the early 2030s, and that timeline is the one that actually matters for your electricity bill.
Helion's different bet
CFS is not alone, and its main rival is chasing a genuinely different design. Helion is not building a tokamak; it uses a pulsed, linear machine that aims to convert fusion energy to electricity directly. In February 2026 it said its Polaris prototype reached about 150 million degrees Celsius and achieved fusion using deuterium and tritium fuel. Helion has an unusually concrete commercial commitment hanging over it: a deal to supply power to Microsoft by 2028, an aggressive date that makes it either the boldest or the most over-promised bet in the field. Two serious companies, two different physics approaches, both now racing against calendar dates rather than vague horizons.
Why it matters
The reason fusion suddenly attracts billions is the same reason it matters: the world, and specifically the AI-and-datacenter build-out, needs enormous amounts of clean, always-on power, and the appetite is growing faster than the grid can comfortably supply. Fusion is the biggest version of that promise: abundant fuel (deuterium from seawater, though tritium has to be bred from lithium), no carbon, no meltdown risk, no long-lived waste. If any of these companies delivers even a working demonstration plant in the 2030s, it reshapes the century.
So hold both thoughts at once. The progress is real, the money is real, and for the first time the leaders are confident enough to name years, which they would not do if they expected to fail publicly. And every one of those years is still a target, "net gain" is not yet "net power," and fusion has humbled confident people before. The honest verdict is that fusion has moved from "maybe someday" to "here is the machine, here is the date, watch it," which is the most exciting place it has ever been, and still not the same as done. For more, see the Future section and our look at the nuclear reactors AI is reviving.
The fusion race, at a glance
| Commonwealth Fusion (CFS) | ~$4B raised (incl. $1B in July 2026); SPARC tokamak ~80% built |
| CFS targets | First plasma later in 2026; net energy gain (Q>1) in 2027; ARC power plant early 2030s |
| The enabling tech | ~20-tesla high-temperature superconducting (HTS) magnets |
| Helion | Different pulsed design; Polaris hit ~150M°C; deal to power Microsoft by 2028 |
| The key caveat | "Net gain" in the plasma ≠ net electricity to the grid |
| Why now | AI/datacenter demand for clean, always-on power |
Four new American reactors just 'went critical'. Here's what that actually means
The US government set a patriotic deadline: get new nuclear reactors running by 4 July 2026. Four made it, beating the goal, in some of the fastest reactor builds in decades (a fifth followed in August). It is a genuine milestone driven by AI's hunger for power. It is also not what the headlines imply: not one of them is yet putting a single watt onto the grid.
Quantum computing just had its 'this is a real industry now' week
In a few days in August, quantum computing companies posted triple-digit revenue growth, went live inside a big cloud platform, and landed government chip money. It is a genuine inflection, from lab curiosity to business. It is also not the same as quantum computers being broadly useful yet, and telling those two apart is the whole point.
Google's Pixel 11 is here, and the real product is Gemini everywhere
At Made by Google 2026, Google unveiled a full Pixel 11 line, a folding flagship, a new watch and its first Bluetooth tracker, all built on a 2nm Tensor G6 chip. But the hardware was the supporting act. The headline features are AI ones, a camera that turns sign language into text and a voice input that understands rambling speech, and the prices crept up about a hundred dollars. Here is what launched and what it signals.