General Fusion Hits 1.1 keV With the UK's Fusion Lab. What the Milestone Does and Doesn't Prove

General Fusion and the U.K. Atomic Energy Authority (UKAEA) announced on Oct. 8, 2026 that the company's LM26 machine reached an electron temperature of 1.1 keV, about 12.6 million degrees Celsius, as measured by a new diagnostic the two organizations built together. It is a real step for a company pursuing an unusual approach. It is also a narrower claim than the headlines suggest, and the underlying paper has not yet been through peer review.

What Was Announced

• The number: 1.1 keV of electron heating, which the release converts to about 12.6 million °C. A straight conversion of 1.1 keV gives roughly 12.8 million °C, so the small gap is likely rounding in the underlying measurement. We could not read the paper to confirm.
• The method: Thomson scattering, which fires a laser into the plasma and reads the light that scatters off electrons. The reading was taken just before peak compression.
• The machine: Lawson Machine 26 (LM26), which uses Magnetized Target Fusion. A magnetized plasma is squeezed mechanically over milliseconds, rather than held in place by magnets or blasted by lasers.
• The claim of a first: the companies say it is the first fusion machine to reach 1 keV using low-speed compression. That is the companies' framing, not an independent ranking.
• The UK role: UKAEA co-authored the paper, helped design and commission the Thomson scattering diagnostic, and built its polychromator at its DICE facility at Culham.

Where It Sits in the Roadmap

This follows General Fusion's June 23, 2026 announcement that LM26 had reached about 8.4 million °C (0.72 keV, with an uncertainty of ±0.08 keV), heated only by compression after the plasma was formed. In that release, 1 keV was named as the next major target. The company's stated path from here is a 10 keV milestone, roughly 100 million °C, and then the Lawson criterion, the combination of density, temperature and confinement time needed for net fusion energy. Upgrades to LM26 for higher compression ratios are underway, and the company says it aims to operate a first-of-a-kind plant around 2035.

"Hot electrons are a necessary step for proving fusion confinement concepts." — Prof. Dennis Whyte, UKAEA CEO, in the joint release

What It Does and Doesn't Prove

This section is analysis, not reporting. Three points are worth holding onto.

• It is an electron temperature. The release reports no ion temperature. In fusion, ions are the particles that fuse, so ion temperature matters at least as much for reactor performance. Electron heating is a necessary step, as Whyte says, not a sufficient one.
• It is not peer reviewed yet. The paper has been posted on the company's site and submitted for peer review, and UKAEA's CEO says he looks forward to its publication in Review of Scientific Instruments. Until then, treat the figure as a company and partner claim.
• Temperature is one of three ingredients. The Lawson criterion also requires density and confinement time. A plasma that is hot but short-lived or thin does not yield net energy. 1.1 keV is also about a tenth of the 10 keV the company targets next.

Even so, the step from 0.72 keV in June to 1.1 keV in October, measured with a diagnostic that an outside national lab helped build and check, is the sort of incremental, measurable progress the field needs. It also shows how different fusion's bets are. Our guide to fusion milestones in 2026 covers the tokamak route Commonwealth Fusion Systems is taking, and the same week brought news on how that company is using AI to speed it up, which we cover in our piece on the AI digital twin of SPARC.

Why It Matters

• It is a measured step, not a breakthrough to net energy. The result is 1.1 keV electron heating, against a stated next target of 10 keV.
• Independent involvement helps credibility. A national lab co-developed the diagnostic, though peer review is still pending.
• It strengthens a different fusion route. Compression-based Magnetized Target Fusion competes with magnetic-confinement tokamaks and laser approaches.
• The timeline is still the company's. A first plant around 2035 remains a target with stated caveats, not a commitment.