Private Fusion Power Is No Longer a Punchline: Where Commonwealth Fusion, Helion, and TAE Technologies Stand in Mid-2026

For most of the twentieth century, nuclear fusion was a science project that perpetually promised commercial power "in thirty years." The joke aged poorly, but the timeline did not improve — until private capital arrived at scale. Between 2021 and 2025, Commonwealth Fusion Systems (CFS), Helion Energy, and TAE Technologies collectively raised more than $5 billion. Each company is pursuing a distinct technical path, and by mid-2026 the gap between them and the national-lab programs they spun out of has widened considerably. This article examines what each company has actually demonstrated, where engineering risk remains concentrated, and what a credible path to grid electricity looks like for each.

Commonwealth Fusion Systems: The Tokamak Bet on High-Temperature Superconductors

CFS is the closest thing fusion has to a mainstream front-runner. The company was founded in 2018 as a spin-out of MIT's PSFC lab and has raised approximately $1.9 billion, with a $1.8 billion Series B closed in late 2021 anchored by Tiger Global and Khosla Ventures. Its entire strategy rests on one material breakthrough: REBCO high-temperature superconducting (HTS) tape, which allows magnets to operate at far higher field strengths than the low-temperature superconductors used in ITER.

In September 2021, CFS demonstrated a 20-tesla magnet — the world's most powerful superconducting magnet of its kind — using its REBCO coil technology. That number matters because magnetic field strength scales plasma confinement as B-squared, meaning a doubling of field strength quadruples confinement quality. The demo validated the core engineering premise of SPARC, CFS's planned demonstration tokamak scheduled to achieve first plasma in 2025 and net energy Q>2 by 2027.

As of mid-2026, CFS reports that SPARC construction at its Devens, Massachusetts facility is on schedule, with vacuum vessel assembly complete and magnet installation underway. The company has not yet announced a formal net-energy date, but internal project timelines suggest a 2027–2028 window for a Q>1 result. If SPARC succeeds, the follow-on commercial plant — ARC — would be approximately 400 MW and is targeted for grid connection by the mid-2030s. CFS has signed a formal site option agreement with Dominion Energy for a Virginia location.

The remaining engineering risks at CFS are significant but bounded. Plasma physics at the field strengths SPARC will operate under is not fully characterized — disruptions (sudden plasma collapse events) at high field could damage first-wall components faster than predicted. CFS's physics team argues that the small tokamak geometry actually reduces disruption energy, but this remains undemonstrated at SPARC-scale parameters.

Helion Energy: Pulsed Fusion With a Direct Electricity Extraction Claim

Helion's approach is architecturally different from a tokamak. The company uses a field-reversed configuration (FRC) — a compact, elongated plasma that is magnetically self-confined without the external coil complexity of a tokamak. Helion compresses two plasmoids together in a linear accelerator geometry, triggering a fusion pulse, and then claims it can recover electricity directly from the decelerating magnetic field rather than through a thermal steam cycle.

Founded in 2013 and headquartered in Everett, Washington, Helion has raised roughly $570 million, with a $500 million Series E in 2021 led by Sam Altman, who also became its chairman. The company's seventh prototype, Polaris, is under construction and is designed to demonstrate net electricity production — not just scientific Q>1, but actual electricity out exceeding electricity in.

Helion's most publicized milestone was a 2023 announcement that it had reached 100 million degrees Celsius plasma temperature in its sixth prototype (Trenta), exceeding the threshold needed for D-He3 fusion, the fuel cycle Helion ultimately prefers because it produces far fewer neutrons than D-T fusion. In November 2021, Microsoft signed a power purchase agreement with Helion — the first PPA of its kind in fusion history — targeting delivery of at least 50 MW by 2028, with financial penalties if the deadline is missed.

By mid-2026, Helion has not publicly confirmed that Polaris has achieved net electricity. The company maintains that it is on track for a 2028 commercial demonstration, but independent fusion analysts note that going from temperature milestones to net-electricity extraction involves several unproven engineering steps: sustaining the compressed plasma long enough for meaningful energy recovery, demonstrating that the direct-conversion mechanism works at scale, and achieving sufficient pulse repetition rate for a commercial plant. The Microsoft PPA deadline creates unusual financial accountability rarely seen in fusion development.

TAE Technologies: The Boron Path and a 30-Year Persistence

TAE Technologies (formerly Tri Alpha Energy) is the oldest and most contrarian of the three. Founded in 1998 in Foothill Ranch, California, TAE has raised over $1.2 billion and has operated nine prototype machines over 26 years. Its target fuel cycle — proton-boron (p-B11) — is considered by mainstream fusion physics as the hardest possible: it produces almost no neutrons, making it an ideal clean fuel, but it requires plasma temperatures exceeding one billion degrees Celsius and cross-sections far lower than D-T or D-He3.

TAE's current machine, Copernicus, is designed to reach the plasma conditions required to demonstrate p-B11 viability. The company claims to have achieved stable plasma operation at 75 keV (roughly 870 million degrees) in earlier prototypes. By mid-2026, TAE has not announced a net-energy demonstration date, and its public communications have shifted toward near-term revenue from adjacent technologies — particularly a power management and storage business (TAE Power Solutions) that applies its plasma control expertise to EV battery systems.

TAE's physics advisors include Nobel laureate Burton Richter and former CERN director Carlo Rubbia, and the company's approach is taken seriously in the physics community. However, the timeline to commercial p-B11 fusion remains the most speculative of the three. TAE executives have stated they believe a commercial plant is achievable in the 2030s, but no credible independent analysis has confirmed this timeline for p-B11 specifically. The company's longevity and its diversification into adjacent revenue streams suggest a strategy of sustained development rather than a sprint to a single demonstration milestone.

What the Differences Actually Mean for Grid Electricity

The three companies are not simply racing down the same road at different speeds — they are pursuing different roads with different risk profiles.

• CFS/SPARC has the most conventional physics (D-T tokamak) and the most demonstrated hardware milestone (the 20T magnet). The path to ARC is long but the physics is the best understood. If SPARC confirms net energy in 2027–2028, a mid-2030s grid connection is plausible but requires sustained capital and regulatory approvals that do not yet exist at scale.
• Helion has the most concrete commercial accountability via the Microsoft PPA, which creates a 2028 pressure point that will either validate or publicly stress-test its timeline. The direct electricity extraction claim is novel and, if demonstrated, would represent a significant efficiency advantage over thermal cycles. The risk is that Polaris must demonstrate multiple unproven steps simultaneously.
• TAE is pursuing the cleanest long-term fuel cycle at the cost of the hardest near-term physics. Its timeline is the least defined and its near-term revenue pivot suggests the company itself is managing timeline uncertainty. A commercial p-B11 plant before 2040 would require breakthroughs not yet demonstrated at any scale.

The Regulatory and Grid Integration Gap Nobody Talks About

Even if all three companies hit their self-stated physics milestones on schedule, a fusion plant delivering electrons to the U.S. grid faces a licensing pathway that does not yet exist in final form. The Nuclear Regulatory Commission published a proposed fusion regulatory framework in 2023 that would treat fusion facilities more like accelerators than fission reactors — a significant reduction in regulatory burden — but final rules have not been promulgated. Utilities integrating a first-of-kind 400 MW baseload fusion plant will require interconnection studies, insurance frameworks, and operational protocols that will take years to develop in parallel with the physics program.

The honest summary for mid-2026: CFS is the closest to a defined physics milestone with demonstrated hardware, Helion has the most accountable commercial timeline, and TAE is the longest technical bet with the cleanest ultimate fuel cycle. Grid electricity from fusion before 2035 remains unlikely for any of the three. A 2035–2040 window for the first commercial kilowatt-hours is achievable if SPARC or Polaris succeeds and regulatory frameworks advance in parallel. That is not a guarantee, but it is no longer a joke.

What to Watch in the Next 18 Months

• SPARC first plasma announcement from CFS — expected 2025–2026, now likely slipping into late 2026 or early 2027 based on construction pace
• Helion Polaris net-electricity result — any announcement before end of 2026 would significantly de-risk the 2028 Microsoft PPA deadline
• NRC fusion licensing final rule — publication expected in 2026, which will set the regulatory cost structure for commercial plants
• TAE Copernicus plasma temperature milestone — a confirmed 1-billion-degree result would be a major physics validation even without net energy
• New capital raises — the next funding rounds for all three will signal investor confidence in their respective timelines