China’s HL-4 Fusion Tokamak Will Test 25-Tesla Magnets Under Reactor Conditions

China’s new HL-4—or Huanliu-4—is being positioned as an engineering testbed for the next stage of magnetic-confinement fusion. Its immediate mission is to determine whether high-temperature superconducting (HTS) magnets can remain reliable while operating in the combined heat, radiation and mechanical environment of a fusion experiment, rather than to generate commercial electricity. 12

HL-4’s main job: validate magnets in a fusion environment

Tokamaks use powerful magnetic fields to confine plasma. The proposed HL-4 facility is intended to move HTS magnet technology beyond isolated laboratory testing and into conditions that more closely resemble those inside a future fusion reactor.

Conference materials described HL-4 as a fourth-generation magnetic-confinement fusion experimental device and as a platform for systematically verifying the reliability of high-field HTS magnets under “complex fusion conditions.” 810 That distinction matters: a magnet can perform well in a controlled test, yet face a far harder challenge when exposed to intense plasma heat loads, radiation and the large electromagnetic forces associated with high-field operation.

The project therefore addresses a central bottleneck in fusion engineering. Stronger magnets can help make tokamaks more compact and improve plasma confinement, but their materials, structures, cooling systems and control equipment must all continue to work reliably as part of an integrated machine.

HL-4 follows the Southwestern Institute of Physics’ Huanliu series

HL-4 is the planned successor to a sequence of devices developed by China’s Southwestern Institute of Physics (SWIP). That lineage includes HL-1, HL-1M, HL-2A and HL-3, with HL-3 formerly known as HL-2M. The institute’s earlier machines have progressively supported higher-performance plasma research; HL-3 achieved first plasma in 2020 and high-confinement operation in 2023. 2627

HL-3’s 2025 results provide the plasma-research context for the next machine. An official CNNC announcement reported an ion temperature of 117 million °C and an electron temperature of 160 million °C. 19 Other 2026 reporting described the 2025 result as roughly 120 million °C for ions alongside 160 million °C for electrons. 1825 The difference reflects different descriptions of the reported milestone, so 117–120 million °C is the safer range rather than a single uncontested figure.

HL-4’s proposed role is different from simply setting another plasma-temperature record. It is intended to test whether the magnet and reactor systems needed for sustained, high-field operation can tolerate the conditions created by that plasma.

The linked 25-tesla magnet roadmap

A Yangtze River Delta innovation consortium led by China Fusion Energy announced a separate development program for high-field HTS tokamak magnets intended to meet HL-4’s requirements. The consortium includes nine research and industrial organizations, including China Fusion Energy, Shanghai Superconductor, Eastern Superconductor, Shanghai Jiao Tong University and Shanghai Electric Nuclear Power Group. 56

Its roadmap calls for a 25-tesla HTS high-field magnet development and testing line by 2028, followed by prototype magnet development by 2030. 45

The 25-tesla target is important because it turns HL-4 from a single facility announcement into a broader manufacturing and qualification program. Reaching the target will require more than producing superconducting material: the project also needs magnet fabrication, structural support, cryogenic systems, power supplies, testing infrastructure and the ability to integrate those parts into a tokamak environment.

Why China’s 582-ton magnet matters—and what it does not prove

Another Chinese fusion project illustrates the scale of the hardware challenge. In Hefei, researchers completed and tested a 582-ton superconducting coil described as the world’s largest fusion magnet. The D-shaped coil measures about 21 meters by 12 meters and was built for the Burning Plasma Experimental Superconducting Tokamak, or BEST. 113

BEST is being built with a goal of demonstrating net fusion power gain and electricity generation around 2030. 14 The successful testing of its giant magnet is therefore a significant manufacturing and integration milestone: it shows progress in building the components required to confine extremely hot plasma at reactor scale.

It is not, however, evidence that China has already produced a commercial fusion reactor or achieved economical, continuous fusion electricity. BEST’s 2030 objective remains a planned demonstration, just as HL-4 and the 25-tesla magnet remain planned engineering milestones. 14

Fusion is becoming an industrial strategy

Taken together, the HL-4 design, the 25-tesla magnet consortium and the BEST hardware program point to a strategy that extends beyond plasma physics. China is developing capabilities across superconducting materials, magnet manufacturing, large-scale testing and reactor engineering, while linking research institutions with state-backed industrial companies. 59

Fusion has also been identified as a future-technology priority in China’s 15th Five-Year Plan period, through 2030. 33 Other reporting has described plans for more than 300 billion yuan in investment before 2030, although that figure should be treated as a reported policy and industry estimate rather than as a confirmed HL-4 project budget. 37

The strategic objective is straightforward: reduce dependence on foreign equipment and materials, build a domestic fusion supply chain and gain experience in the difficult transition from experimental physics to reactor engineering. That places China’s program in the wider international race to turn fusion research into a practical energy technology.

The bottom line

The Shanghai announcement was less a claim that fusion power has arrived than a statement about the next engineering problem China wants to solve. HL-4 is designed to test whether high-temperature superconducting magnets can operate reliably under realistic fusion-reactor conditions, with a related consortium targeting a 25-tesla prototype by 2030.

HL-3’s high-temperature plasma results, BEST’s 582-ton magnet and the planned HTS manufacturing network show a program moving toward larger-scale engineering. But the decisive milestones—reliable long-duration operation, net power and economically viable electricity—still lie ahead.