Fusion Startup and Company Directory: 42 Companies by Country, Founding Year and Technology

Summary: The fusion industry includes not only power-plant developers, but also companies making magnets, fuels, materials, lasers and neutron-based services. This directory compares 42 companies by country, founding year, approach and business role. Even companies pursuing the same reactor family can be developing different products at different stages.

Reference date: October 30, 2026. This is a comparison based on public primary sources, not an investment recommendation or a census claiming to include every company worldwide.

A list of names alone does not explain an industry. One company designs a power plant; another produces the magnets it needs. Others use neutrons for testing or isotope production. All are connected to fusion, but their products, customers and required demonstrations differ.

“Approach” in this directory mainly means how plasma is confined or compressed, not the ultimate electricity-generator technology. Many concepts investigate heat-based power systems; some aim to recover charged-particle or magnetic energy directly as electricity. A confinement label alone does not establish efficiency, cost or delivery-date superiority.

• Tokamaks: use magnetic fields to confine a toroidal plasma. Spherical tokamaks pursue a geometry with a relatively small central opening.
• Stellarators and helical devices: create twisted magnetic fields mainly with external coils. They combine potential advantages for continuous operation with demanding magnet and structural manufacturing.
• Laser inertial fusion: rapidly compresses and heats small fuel targets. Power generation also requires efficient lasers, mass-produced targets, repeated delivery and durable chamber walls.
• FRCs, mirrors and dipoles: include several magnetic geometries beyond conventional toroidal designs. Fuel, heating and energy-recovery choices differ between companies.
• Z-pinches, magnetized targets and pulsed power: use high currents, mechanical compression or related methods. The path from individual experiments to repetitive operation matters.

Connecting a technology name to a real research machine makes the company table easier to read. JT-60SA, pictured here, is a superconducting tokamak developed jointly by Japan and Europe. It studies high-performance plasma operation, supports ITER and future-reactor research, and trains researchers and engineers. Controlling a toroidal plasma requires heating, diagnostics, vacuum and cooling systems as well as the central machine.

CFS and Tokamak Energy also work on tokamak technologies, but JT-60SA belongs to neither company. Shared principles do not make their geometries, magnets, fuels or operating objectives identical. QST’s research objectives · F4E’s photograph and explanation of auxiliary systems

Wendelstein 7-X (W7-X), pictured here, is a stellarator at Germany’s Max Planck Institute for Plasma Physics (IPP). It studies three-dimensional magnetic confinement produced by external coils and long-duration plasma operation. Its surrounding pipes and diagnostic equipment also show integration work that a simple reactor schematic cannot convey.

Proxima says it spun out of IPP in 2023 to build on W7-X research. W7-X is not a power plant built by Proxima. Turning research knowledge into a commercial plant still involves different magnet designs, manufacturing, maintenance and integration with generating equipment. IPP’s W7-X material · Proxima’s research lineage

Companies are included where public sources identify development of fusion reactors or devices, or fusion-focused products and services. More mature companies and corporate-group ventures are included in separate categories so that the boundaries remain clear. Country refers to principal corporate or operating locations, not founders’ nationality or research partners’ locations. Multiple countries are shown where relevant.

Founding year generally means the company’s formation, not the year its researchers began studying fusion. Unverified dates are not filled in by inference. Where sources differ—as with GenF’s business formation and formal launch, or Xcimer’s founding-year descriptions—the distinction is retained.

The FIA’s 2026 industry survey received responses from 56 companies. That is not the total number of companies in the world: membership, participation and classification affect coverage. This directory likewise states its public-information boundaries. FIA’s explanation of its 2026 survey

The essential comparison is between like-for-like achievements. Energizing a magnet, producing plasma, generating fusion reactions, producing neutrons, achieving plasma or target energy gain, demonstrating net electricity at facility level, and supplying electricity over extended periods are different milestones.

For example, designing a laser power plant using NIF’s target-gain results does not mean the company’s own plant has delivered electricity. Producing neutrons and supplying testing services likewise does not establish the energy balance required for a power station. Power-purchase agreements and development partnerships can be important commercial preparations, but they are not proof that future delivery has occurred.

When reading “the 2030s,” “20 MW” or a number of shots per second, ask whether the number is a measurement, an equipment specification or a design objective. Electron and ion temperatures differ, as do instantaneous power and sustained energy production. This directory does not turn company plans into independently verified achievements.

The spherical vessel in the photograph is the target chamber at LLNL’s NIF research facility, not a power plant owned by a company in this directory. NIF experiments compare fusion yield with laser energy delivered to the target. Longview describes drawing on its team’s experience designing, building and operating NIF. This is a concrete example of why a research facility’s accumulated knowledge must be distinguished from a company’s future generating plant. LLNL on ignition · Longview’s company description

Japan is not pursuing only one reactor family. Helical, laser, advanced-fuel and HTS tokamak development sits alongside fusion engineering and materials supply. Separating reactor developers from equipment and materials providers helps explain why the same supplier can appear in projects in several countries.

For manufacturers considering participation, today’s demonstrator requirements can be more concrete than a predicted date for a completed power plant. Magnets involve conductors, winding, joints, cryogenics and inspection; laser systems involve optics, power supplies, targets and positioning; blankets involve materials, fluids, corrosion and fuel recovery. This is an editorial decomposition of technical work, not confirmation that a particular company is currently procuring those items externally.

To understand a company in depth: distinguish its experimental device from its future power plant; read papers as well as company announcements; examine the fuel, heat, maintenance and regulatory systems outside the plasma; and distinguish deliveries from planned collaborations.

A young company does not necessarily have a young research base. Many bring decades of university and national-laboratory work into a business. Conversely, a long history or a large financing round does not prove net electricity or commercial economics. Start by asking who is trying to deliver what, and at which stage. This directory is an entry point for making that comparison.