China accounts for around 91% of refined rare earth oxide output and 92% of global rare earth permanent magnet production, making its processing and manufacturing capacity a decisive constraint on magnet supply [S4][S5].
Global dependence is narrower than China’s 60% share of mined magnet rare earths in 2024, because separation and refining turn ore into the oxides, metals and alloys required for NdFeB magnets [S2].
Separation and Processing Capacity Is Concentrated in China
China accounted for 60% of global mined production of magnet rare earths in 2024, but its share rises sharply in the processing stages that determine usable material specifications [S2]. A separate October 2025 assessment placed China at approximately 70% of rare earth mining, 90% of separation and processing, and 93% of magnet manufacturing [S5].
The figures measure different links in one value chain, not conflicting substitutes. Mining supplies feedstock, separation divides mixed rare earths into individual elements, refining and metallisation produce metals and alloys, and magnet manufacturing converts those inputs into finished components [S3]. A new mine does not remove the bottleneck if the separated oxide cannot be produced elsewhere at commercial scale.
Concentration also extends beyond separation. China holds around 91% of refined rare earth oxide output and 92% of global rare earth permanent magnet production [S4]. Magnet-critical elements such as neodymium, praseodymium, dysprosium and terbium have strategic value because of their use in permanent magnets [S3].
Magnet Supply Has Four Separate Chokepoints
Rare earth permanent magnets serve electric vehicles, wind turbines, robotics, industrial motors, consumer electronics, aerospace and defence systems, so disruption can propagate across several manufacturing sectors [S4]. Their production sequence requires mining feedstock, separated oxides, refined metals or alloys, magnet manufacturing and, where applicable, export approval.
China’s share was approximately 95% of global NdFeB output in a December 2025 supply-chain assessment [S8]. The range of 92% to 95% reflects different datasets and definitions, including global magnet output versus global NdFeB output, but both estimates place the manufacturing stage firmly within China’s sphere of dominance [S4][S8].
That concentration explains why Europe’s dependency extends beyond imported ore. The European Union has virtually no domestic rare earth production and sources 98% of its rare earth magnet demand from Chinese suppliers [S7]. For buyers, a signed mining contract does not by itself guarantee delivery of a qualified magnet.
Separation Capacity Is Harder to Replace Than Mining
Alternative supply chains remain capital-intensive and technically demanding because competitive resilience requires separation, processing, downstream manufacturing and recycling rather than mining capacity alone [S3]. The United States has limited domestic production capacity for sintered neodymium magnets used in wind turbines and electric vehicles [S1].
Capability in rare earth separation, refining, and permanent magnet manufacturing is heavily concentrated, and China’s central position in the value chain reflects decades of coordinated industrial development, technological learning, and vertical integration [S3]. These constraints explain why mine announcements do not translate directly into separated oxide output. Solvent extraction, selective separation, purification and metallisation must operate at consistent yield, quality and cost targets.
European policy also links diversification to the Critical Raw Materials Act, but the practical constraint remains the full processing chain [S3]. A mine, separated oxide plant, metal production line and magnet factory can face different commissioning schedules, so each project must be assessed as a separate capacity addition rather than as one complete supply source.
Export Licensing Can Turn Capacity Concentration Into Acute Shortage
China’s Ministry of Commerce Announcement No. 61 of 2025 expanded rare earth and permanent magnet export controls. Under the new controls, foreign firms must obtain Chinese government approval to export magnets [S5].
In 2025, restrictions on rare earth oxides, processing technologies and magnet manufacturing contributed to manufacturing disruptions as magnet supplies ran out [S4].
The failure mode is therefore not simply a physical shortage of mined material. An installed separation line, metal conversion operation or magnet plant may be unable to ship because the product, its input or its process technology falls within an export licensing regime. Buyers need traceability and process provenance in addition to reserve inventories.
Mine Supply, Separation and Magnet Output Must Be Compared Separately
China holds the leading position in each major stage, but the percentages are not interchangeable. China accounted for 60% of global mined magnet rare earth production in 2024, around 91% of refined rare earth oxide output, and 92% of global rare earth permanent magnet production, while its share of global NdFeB output was around 95%.
On feedstock concentration, China leads at about 60% to 70% [S2][S5]. China accounts for around 91% of refined rare earth oxide output. On magnet manufacturing, published estimates range from 92% for global rare earth permanent magnet production to 93% in the October 2025 assessment and roughly 95% for global NdFeB output [S4][S5][S8]. This progression shows that dependency increases after mining.
For procurement decisions, the practical conclusion is clear: qualify the full route from designated mine to separated oxide, metal or alloy, and finished magnet. A supplier offering attractive mining exposure may still depend on Chinese separation or magnet conversion. NdFeB magnet price breakdown: rare earth content becomes more useful when reviewed alongside processing provenance rather than ore price alone.
Volume Buyers, Mission-Critical Users and Investors Face Different Risks
Large-volume manufacturers benefit most from dual sourcing because a qualified alternative can reduce exposure to one country’s separation and export decisions. Europe’s 98% Chinese sourcing rate for rare earth magnet demand indicates a concentrated buyer position that new processing projects must address [S7]. Defence, aerospace and semiconductor users face a stricter risk profile because export licences can determine whether material reaches the intended programme [S5].
Investors and project developers face a different constraint: capacity without dependable offtake may underperform commercially. A proposed economic security agreement would use capacity-linked market access and qualified supply to create demand for non-Chinese oxides, metals, alloys and magnets, with a G7 definition identifying excessive single-supplier market share as 60% or more across the rare earth permanent magnet supply chain [S4].
This approach is not a universal fit for every buyer. Small projects without long-term demand may struggle to finance separation facilities, while firms that require identical magnet chemistry face qualification and testing constraints [S3][S4]. Demand-side aggregation is more relevant to strategic users than to consumers purchasing small quantities.
Traceable Non-Chinese Capacity Is the Measurable Supply-Security Test
The strongest near-term resilience signal is qualified separation and magnet output from outside China, not another mining announcement. Project screening should verify separated-oxide nameplate capacity, actual commissioning, process ownership and technology dependencies, magnet-line qualification and contracted output rather than relying on resource estimates alone [S3][S4].
Demand-side measures add another measurable signal. Contracts covering qualified non-Chinese oxide, metal, alloy and magnet volumes would distinguish bankable diversification from subsidised plants without secured offtake [S4]. Buyers can also track the ratio of separated material available outside China to total magnet demand, including whether alternative projects have reached commercial production rather than pilot status [S3].
Humanoid robots are one potential demand source: the projected addition of 449 kt of NdFeB magnet demand would increase pressure on an already concentrated chain [S1]. That scale makes the processing bottleneck more consequential, not less, because higher magnet demand requires more separated neodymium and praseodymium feedstock as well as finished magnet capacity.