$MP 20261113 35.0 PUT$ The U.S. needs roughly an additional 18,000 metric tonnes per year of usable rare-earth compounds/metals to cover its current aggregate import gap—but that is only a broad quantity benchmark, not full independence from China. The much harder requirement is targeted capacity for separated NdPr and especially heavy rare earths such as dysprosium and terbium, plus metal/alloy and permanent-magnet conversion.

Available 2025 data indicate U.S. apparent rare-earth consumption of about 27,000 tonnes , against domestic compounds-and-metals output of roughly 8,900 tonnes . The implied gap is 18,100 tonnes per year (27,000 − 8,900 = 18,100). Net import dependence was reported at 67% .

Xueqiu

What “enough capacity” means

Supply-chain stage Capacity the U.S. needs Why it matters

Mine/concentrate Not sufficient on its own The U.S. already has meaningful ore production, but ore does not replace separated material.

Separation/refining About 18,000 tpa across the relevant rare-earth products, if the aim is to replace current net imports This is the central bottleneck: separation turns concentrate into individual, high-purity oxides.

Heavy rare-earth separation A new, dedicated commercial capability The U.S. reportedly remained 100% net-import dependent for strategic heavy rare earths; aggregate tonnage can therefore be misleading.

Xueqiu

NdPr metal/alloy and magnets Capacity aligned with domestic EV, wind, industrial and defence demand These steps determine whether domestic oxide production actually translates into reduced China exposure for high-performance magnets.

The practical answer is smaller for diversification, larger for independence

To materially reduce reliance: A credible first objective is not replacing every imported rare-earth product, but adding reliable non-China separation and magnet capacity for the highest-risk products—NdPr, dysprosium and terbium. This can cut concentrated supply-chain risk even before total U.S. capacity reaches 18,000 tpa.

To replace today’s net imports onshore: The benchmark is approximately 18,100 tpa of output, assuming the reported consumption and domestic-output definitions are comparable.

To become resilient rather than merely self-sufficient: Capacity must exceed average demand enough to accommodate outages, quality specifications, recycling losses and demand growth. That requires multiple qualified plants and non-China feedstock, not one domestic facility.

[Symbol:MP|MP Materials (MP)] is strategically relevant because its California separation operation and Texas magnet plans address the missing midstream and downstream links. A reported target for its Texas magnet facility is 10,000 tonnes by 2028 , which would be substantial relative to the current U.S. processing gap, but magnet tonnes cannot be directly netted against rare-earth oxide/metal tonnes: they are different products with different mass balances.

Xueqiu

Key limitation: Current data do not provide a clean, audited product-by-product U.S. demand breakdown for NdPr, dysprosium and terbium, nor verified operating capacity at every U.S. separation facility. Therefore, 18,100 tpa should be treated as an aggregate import-replacement benchmark, not a precise capacity-build target for individual elements.

The main factors that could reduce the needed domestic build are supply agreements with Australia, Brazil or other non-China sources, recycling, and lower magnet intensity per end product. The factors that could raise it are faster EV/robotics/defence demand growth and persistent lack of heavy-rare-earth processing.

Information is for reference only and does not constitute investment advice.