Fusion: where the field actually stands

Fusion is the tracker's flagship field, and it currently carries six metrics, 42 milestones and the widest gap on the site between what has been demonstrated and what has been financed.

Start with the headline numbers. NIF holds the energy-gain record at Q 4.1 (April 2025), measured as fusion yield against the laser energy delivered to the target — the eighth ignition shot, and more than quadruple the light that reached the fuel. Peak temperature sits at 150 million °C, set by Helion's Polaris in February 2026, the first privately built machine to do deuterium-tritium fusion; EAST's 160 million °C from 2021 remains the higher figure on that axis. The longest sustained plasma is WEST's 1,337 seconds — about 22 minutes — from February 2025, roughly 25% past EAST's 1,066 s. Private capital into the field totals about $9.77B, and published fusion research runs a little over 1,100 papers a year.

Now the number that complicates the story. The triple product — density times temperature times confinement time — is the fairest single axis for comparing confinement approaches, and breakeven needs roughly 10²¹. The record is 1.5×10²¹, set by JT-60U in August 1996. No magnetic-confinement machine has beaten a thirty-year-old result on the measure that matters most. Records on temperature and duration are real, but they are records on one axis at a time, often at different performance levels; a long pulse at moderate temperature and a brief flash at an extreme one are not the same achievement.

What moved in 2026 was mostly institutional rather than physical. Helion secured the world's first full regulatory licences for a fusion power plant in June; Tennessee issued the first licence anywhere for a commercial fusion machine on 31 August, covering Type One Energy's 400 MW stellarator at Bull Run. Pacific Fusion broke ground on a roughly $1B campus in Albuquerque in August. General Fusion became the first publicly traded pure-play fusion company in July. Helion raised $465M at a $15.5B valuation, Focused Energy closed the industry's largest Series A at $240M, and Germany joined a multi-billion-euro EU fusion IPCEI. ITER completed its Central Solenoid, the most powerful pulsed magnet ever built.

Read that list carefully: it is licensing, construction and capital, not net energy. A licence is permission to build; groundbreaking is a construction start. No stellarator has produced net energy, and no Pacific Fusion machine has yet produced any fusion at all.

The honest near-term test is Commonwealth Fusion Systems' SPARC, targeting Q above 1 in 2027 as the first commercially relevant machine to try. ITER's first plasma is now 2034 under the rebaselined schedule, with deuterium-tritium operation in 2039. Every figure above links to its primary source on the metric pages; where a number is company-measured rather than peer-reviewed, we label it.