Stellarators vs. Tokamaks: The 2025 Fusion Race Heats Up

Nuclear fusion—the process that powers the Sun—has been "30 years away" for the past 70 years. But in 2025, the landscape has shifted in ways that make this persistent skepticism increasingly difficult to sustain. Private fusion companies have raised billions of dollars. Government megaprojects (ITER, EAST, Wendelstein 7-X) are producing meaningful physics results. And crucially, the competition between two fundamentally different magnetic confinement approaches—tokamaks and stellarators—is generating innovation at a pace that neither approach alone would achieve.

The core physics challenge is the same for both: confine a plasma of hydrogen isotopes (deuterium and tritium) at temperatures exceeding 100 million degrees Celsius, long enough for fusion reactions to produce more energy than is consumed maintaining the plasma. The magnetic field geometries that accomplish this confinement differ dramatically between tokamaks (axisymmetric, with plasma current providing part of the confining field) and stellarators (non-axisymmetric, with the entire confining field generated by external coils).

Thea Energy's Eos: Stellarators Go Commercial

Swanson et al. (published in Nuclear Fusion) describe the design of Eos, a sub-breakeven stellarator that Thea Energy plans to build as a stepping stone to a fusion pilot plant. Eos is not designed to produce net energy; it is designed to produce neutrons from deuterium-deuterium fusion reactions, which will be used to generate tritium and other valuable radioisotopes.

The business logic is pragmatic: rather than waiting for breakeven fusion (which requires solving numerous engineering challenges simultaneously), Eos generates revenue from isotope production while validating the stellarator physics needed for the eventual pilot plant. The physics validation includes:

• Plasma confinement at parameters relevant to energy production
• Fast-ion confinement (critical for the alpha-particle heating that sustains a burning plasma)
• Steady-state operation (stellarators' key advantage over tokamaks)

Carbajal et al. analyze alpha-particle confinement in the Infinity Two stellarator—a quasi-isodynamic, four-field-period design developed by Type One Energy (a separate company from Thea Energy) as a fusion pilot plant targeting 800 MW of fusion power. Alpha particles (helium nuclei produced by fusion reactions) must remain confined long enough to transfer their energy to the plasma; if they escape prematurely, the plasma cannot sustain itself. Their analysis, using Monte-Carlo codes SIMPLE, ASCOT5, and KORC-T, shows favorable confinement properties with no unstable Alfvén eigenmodes at the envisioned 800 MW operating scenario.

Wendelstein 7-X: The Scientific Stellarator

Klinger's overview of Wendelstein 7-X (W7-X)—the world's largest and most advanced stellarator, located in Greifswald, Germany—documents continued progress toward the high-performance plasmas that stellarators must demonstrate to compete with tokamaks.

W7-X was designed to prove that stellarators can achieve energy confinement times comparable to tokamaks of similar size. The device's carefully optimized magnetic field geometry—computed through decades of numerical optimization—is intended to minimize particle and energy losses while maintaining the intrinsic steady-state capability that gives stellarators their primary advantage.

The 2025 progress includes higher plasma temperatures, longer discharge durations, and improved understanding of the plasma-wall interaction that will determine the feasibility of long-pulse operation in future devices.

EAST: Tokamak Steady-State Progress

On the tokamak side, Yu et al. report progress on the EAST tokamak in Hefei, China—achieving I-mode plasma (an improved confinement regime) with divertor detachment. This is significant because:

• I-mode provides the high energy confinement needed for fusion without the edge instabilities (ELMs) that damage plasma-facing materials in the more common H-mode
• Divertor detachment reduces the heat load on the divertor target—a critical engineering requirement for any reactor-relevant tokamak

The combination of I-mode confinement with divertor protection has been a long-sought goal for tokamak research. EAST's demonstration—in steady-state rather than transient operation—brings tokamaks closer to the continuous operation that fusion power plants will require.

Hybrid Approaches: The Best of Both Worlds?

Li et al. (2026) explore a tokamak-stellarator hybrid on the J-TEXT device, using external stellarator-like coils to supplement the tokamak's plasma current with an external rotational transform. The hybrid approach aims to combine tokamak advantages (high plasma pressure, well-understood physics) with stellarator advantages (reduced disruption risk, steady-state capability).

Their result—suppression of tearing modes (a dangerous plasma instability) through the external rotational transform—suggests that the hybrid approach can mitigate one of tokamaks' most serious operational risks: disruptions that dump the plasma's stored energy into the machine wall in milliseconds.

Claims and Evidence

<

Claim	Evidence	Verdict
Stellarators can achieve energy-relevant plasma parameters	Wendelstein 7-X progress; Eos design analysis	✅ Supported (approaching)
Stellarators offer intrinsic steady-state operation	No plasma current to maintain; fundamental advantage	✅ Physics fact
Tokamaks can achieve I-mode with divertor detachment	EAST demonstration in steady state	✅ Demonstrated
Hybrid configurations mitigate tokamak disruption risk	J-TEXT tearing mode suppression	✅ Supported (proof of concept)
Commercial fusion power is imminent	Major engineering challenges remain	⚠️ Decades, not years

Open Questions

Materials: No material currently exists that can withstand the neutron flux of a fusion reactor for its operational lifetime. Fusion materials research is arguably the pacing item for commercial fusion, regardless of which confinement approach succeeds.

Tritium fuel cycle: Fusion reactors will breed tritium from lithium in the blanket surrounding the plasma. This tritium breeding cycle has never been demonstrated at reactor scale. Can it produce enough tritium to sustain the reactor while also providing startup fuel for new reactors?

Economic competitiveness: Even if fusion works technically, can it produce electricity at a cost competitive with advanced fission, solar, wind, and storage? The economics of fusion remain highly uncertain.

Stellarator magnet technology: Stellarators require magnets with complex 3D shapes—much harder to manufacture than tokamaks' relatively simple coils. Can high-temperature superconducting magnets be fabricated in the complex geometries that optimized stellarators require?

Regulatory pathway: How will fusion power plants be regulated? As nuclear facilities (like fission plants) or under a new regulatory framework that reflects fusion's different risk profile (no meltdown risk, minimal long-lived waste)?

What This Means for Your Research

For plasma physicists, the 2025 results from multiple devices provide a rich dataset for validating computational models of plasma behavior—models that must be reliable before committing billions to fusion pilot plants.

For materials scientists, fusion presents extreme materials challenges (14 MeV neutron irradiation, plasma-surface interaction, high-temperature operation) that drive innovation applicable to other high-performance applications.

For energy policy researchers, fusion's timeline uncertainty (confident that it will work; uncertain about when) creates planning challenges. Should energy policy incorporate fusion into long-term decarbonization scenarios, or treat it as speculative?

Nuclear fusion—the process that powers the Sun—has been "30 years away" for the past 70 years. But in 2025, the landscape has shifted in ways that make this persistent skepticism increasingly difficult to sustain. Private fusion companies have raised billions of dollars. Government megaprojects (ITER, EAST, Wendelstein 7-X) are producing meaningful physics results. And crucially, the competition between two fundamentally different magnetic confinement approaches—tokamaks and stellarators—is generating innovation at a pace that neither approach alone would achieve.

The core physics challenge is the same for both: confine a plasma of hydrogen isotopes (deuterium and tritium) at temperatures exceeding 100 million degrees Celsius, long enough for fusion reactions to produce more energy than is consumed maintaining the plasma. The magnetic field geometries that accomplish this confinement differ dramatically between tokamaks (axisymmetric, with plasma current providing part of the confining field) and stellarators (non-axisymmetric, with the entire confining field generated by external coils).

Thea Energy's Eos: Stellarators Go Commercial

Swanson et al. (published in Nuclear Fusion) describe the design of Eos, a sub-breakeven stellarator that Thea Energy plans to build as a stepping stone to a fusion pilot plant. Eos is not designed to produce net energy; it is designed to produce neutrons from deuterium-deuterium fusion reactions, which will be used to generate tritium and other valuable radioisotopes.

The business logic is pragmatic: rather than waiting for breakeven fusion (which requires solving numerous engineering challenges simultaneously), Eos generates revenue from isotope production while validating the stellarator physics needed for the eventual pilot plant. The physics validation includes:

• Plasma confinement at parameters relevant to energy production
• Fast-ion confinement (critical for the alpha-particle heating that sustains a burning plasma)
• Steady-state operation (stellarators' key advantage over tokamaks)

Carbajal et al. analyze alpha-particle confinement in the Infinity Two stellarator—a quasi-isodynamic, four-field-period design developed by Type One Energy (a separate company from Thea Energy) as a fusion pilot plant targeting 800 MW of fusion power. Alpha particles (helium nuclei produced by fusion reactions) must remain confined long enough to transfer their energy to the plasma; if they escape prematurely, the plasma cannot sustain itself. Their analysis, using Monte-Carlo codes SIMPLE, ASCOT5, and KORC-T, shows favorable confinement properties with no unstable Alfvén eigenmodes at the envisioned 800 MW operating scenario.

Wendelstein 7-X: The Scientific Stellarator

Klinger's overview of Wendelstein 7-X (W7-X)—the world's largest and most advanced stellarator, located in Greifswald, Germany—documents continued progress toward the high-performance plasmas that stellarators must demonstrate to compete with tokamaks.

W7-X was designed to prove that stellarators can achieve energy confinement times comparable to tokamaks of similar size. The device's carefully optimized magnetic field geometry—computed through decades of numerical optimization—is intended to minimize particle and energy losses while maintaining the intrinsic steady-state capability that gives stellarators their primary advantage.

The 2025 progress includes higher plasma temperatures, longer discharge durations, and improved understanding of the plasma-wall interaction that will determine the feasibility of long-pulse operation in future devices.

EAST: Tokamak Steady-State Progress

On the tokamak side, Yu et al. report progress on the EAST tokamak in Hefei, China—achieving I-mode plasma (an improved confinement regime) with divertor detachment. This is significant because:

• I-mode provides the high energy confinement needed for fusion without the edge instabilities (ELMs) that damage plasma-facing materials in the more common H-mode
• Divertor detachment reduces the heat load on the divertor target—a critical engineering requirement for any reactor-relevant tokamak

The combination of I-mode confinement with divertor protection has been a long-sought goal for tokamak research. EAST's demonstration—in steady-state rather than transient operation—brings tokamaks closer to the continuous operation that fusion power plants will require.

Hybrid Approaches: The Best of Both Worlds?

Li et al. (2026) explore a tokamak-stellarator hybrid on the J-TEXT device, using external stellarator-like coils to supplement the tokamak's plasma current with an external rotational transform. The hybrid approach aims to combine tokamak advantages (high plasma pressure, well-understood physics) with stellarator advantages (reduced disruption risk, steady-state capability).

Their result—suppression of tearing modes (a dangerous plasma instability) through the external rotational transform—suggests that the hybrid approach can mitigate one of tokamaks' most serious operational risks: disruptions that dump the plasma's stored energy into the machine wall in milliseconds.

Claims and Evidence

<

Claim	Evidence	Verdict
Stellarators can achieve energy-relevant plasma parameters	Wendelstein 7-X progress; Eos design analysis	✅ Supported (approaching)
Stellarators offer intrinsic steady-state operation	No plasma current to maintain; fundamental advantage	✅ Physics fact
Tokamaks can achieve I-mode with divertor detachment	EAST demonstration in steady state	✅ Demonstrated
Hybrid configurations mitigate tokamak disruption risk	J-TEXT tearing mode suppression	✅ Supported (proof of concept)
Commercial fusion power is imminent	Major engineering challenges remain	⚠️ Decades, not years

Open Questions

Materials: No material currently exists that can withstand the neutron flux of a fusion reactor for its operational lifetime. Fusion materials research is arguably the pacing item for commercial fusion, regardless of which confinement approach succeeds.

Tritium fuel cycle: Fusion reactors will breed tritium from lithium in the blanket surrounding the plasma. This tritium breeding cycle has never been demonstrated at reactor scale. Can it produce enough tritium to sustain the reactor while also providing startup fuel for new reactors?

Economic competitiveness: Even if fusion works technically, can it produce electricity at a cost competitive with advanced fission, solar, wind, and storage? The economics of fusion remain highly uncertain.

Stellarator magnet technology: Stellarators require magnets with complex 3D shapes—much harder to manufacture than tokamaks' relatively simple coils. Can high-temperature superconducting magnets be fabricated in the complex geometries that optimized stellarators require?

Regulatory pathway: How will fusion power plants be regulated? As nuclear facilities (like fission plants) or under a new regulatory framework that reflects fusion's different risk profile (no meltdown risk, minimal long-lived waste)?

What This Means for Your Research

For plasma physicists, the 2025 results from multiple devices provide a rich dataset for validating computational models of plasma behavior—models that must be reliable before committing billions to fusion pilot plants.

For materials scientists, fusion presents extreme materials challenges (14 MeV neutron irradiation, plasma-surface interaction, high-temperature operation) that drive innovation applicable to other high-performance applications.

For energy policy researchers, fusion's timeline uncertainty (confident that it will work; uncertain about when) creates planning challenges. Should energy policy incorporate fusion into long-term decarbonization scenarios, or treat it as speculative?