For years, humanity has aimed to replicate the immense energy of the stars to produce electricity on Earth. Yet, this ambitious goal has consistently appeared to be just a decade away.
Today, a wave of startups is making significant progress, rapidly advancing towards constructing fusion reactors that could supply power to the grid.
These fusion startups have attracted over $10 billion in investments, with more than a dozen securing upwards of $100 million each. In the past year, numerous major funding rounds have concluded, driven by investors interested in the sector as energy needs from data centers increase and fusion technology nears a breakthrough.
The essence of fusion power lies in harnessing the energy released during the fusion of atoms to generate electricity. For decades, humans have had the knowledge to fuse atoms, as seen in the hydrogen bomb, an example of uncontrolled nuclear fusion, as well as various fusion devices developed in laboratories globally. Experimental devices have achieved controlled nuclear fusion, with one even producing more energy than needed to initiate the reaction.
However, none have yet generated a sufficient surplus to sustain a power plant.
To address this challenge, fusion startups are exploring diverse methods. Experts hold differing views on which approaches might succeed, acknowledging the industry’s nascent stage and the absence of guarantees.
The following provides a concise summary of the primary approaches to fusion power.
Techcrunch event
San Francisco, CA
|
October 13-15, 2026
Magnetic confinement
Magnetic confinement is a prevalent technique that utilizes strong magnetic fields to contain plasma, the superheated mixture of particles central to a fusion device.
The magnets required are extremely powerful. For instance, Commonwealth Fusion Systems (CFS) is creating magnets capable of generating 20 tesla magnetic fields, 13 times stronger than a standard MRI machine. These magnets are constructed from high-temperature superconductors, which still need to be chilled to –253˚ C (–423˚ F) with liquid helium.
CFS is currently developing a demonstration device named Sparc on an expedited schedule in Massachusetts. If successful, they plan to start building Arc, their commercial-scale power plant, in Virginia by 2027 or 2028.
Fusion devices employing magnetic confinement come in two main forms: tokamaks and stellarators.
Tokamaks, first conceptualized by Soviet scientists in the 1950s, have been extensively studied since. They come in two basic shapes: a doughnut with a D-shaped profile and a sphere with a central hole. Notable examples include the Joint European Torus (JET), which operated in the UK from 1983 to 2023, and ITER, expected to begin operations in France in the late 2030s.
Tokamak Energy in the UK is developing a spherical tokamak design, with its ST40 experimental machine currently being upgraded.
Stellarators, the other main type of magnetic confinement device, contain plasma within a doughnut-like shape but feature twists and turns unlike the geometric sides of tokamaks. The shape is tailored based on modeling plasma behavior and adjusting the magnetic field to align with its characteristics.
The Wendelstein 7-X, a large stellarator with modular superconducting coils, has been operational in Germany since 2015, managed by the Max Planck Institute for Plasma Physics. Several startups are also working on stellarator designs, such as Proxima Fusion, Renaissance Fusion, Thea Energy, and Type One Energy.
Inertial confinement
Another major fusion approach is inertial confinement, which compresses fuel pellets until the atoms within merge.
Most designs employ pulses of laser light for compression, with multiple laser beams firing simultaneously to converge on the fuel pellet from all directions.
Thus far, inertial confinement is the sole method to surpass the scientific breakeven point, where the reaction yields more energy than it consumes. These experiments were conducted at the National Ignition Facility (NIF) at Lawrence Livermore National Laboratory in California. However, the measurements exclude factors like the electricity needed to power the facility.
Nevertheless, nearly a dozen startups find promise in inertial confinement and are designing reactors using this method. Notable examples utilizing lasers include Focused Energy, Inertia Enterprises, Marvel Fusion, and Xcimer.
Two companies are exploring non-laser alternatives: First Light Fusion, which suggests using pistons, and Pacific Fusion, which intends to employ electromagnetic pulses.
More to come
These are the two principal approaches to fusion power, though not the only ones. Additional details on alternative designs such as magnetized target fusion, magnetic-electrostatic confinement, and muon-catalyzed fusion will be provided soon.
