Capacity tracker
US rare earth capacity tracker: NdPr, heavy rare earths and magnets vs demand
US plants that run today state capacity for about 7,610 t/yr of NdFeB magnets and separate about 3,600 t/yr of NdPr oxide, against 2030 demand of about 38,700 t/yr of magnets and 12,500 t/yr of NdPr oxide. No US plant separates dysprosium or terbium at commercial scale yet; 120 t/yr of dysprosium oxide is under construction.
If every plan were built, magnet capacity would reach about 31,951 t/yr (83% of demand) and NdPr oxide about 10,840 t/yr (87%). Every figure is capacity a company states, not output, and tonnes at different steps are not comparable.
Verified through Oct 1, 2026
Share of 2030 demand
- NdFeB magnets 20% now83% if all built
- NdPr oxide 29% now87% if all built
- Dysprosium oxide 0% now13% if all built
OperatingAll stated plans
- Magnets, operating plants
- 7,610 t/yr20% of 2030 demand; stated, not output
- NdPr oxide, operating
- 3,600 t/yr29% of 2030 demand
- Dysprosium oxide, operating
- 0 t/yr120 t/yr under construction
- Magnets if every plan is built
- 31,951 t/yr83% of 2030 demand
- Rows counted
- 2016 more listed, not added
Capacity against demand
How much rare earth processing and magnet capacity does the US have, and how much is planned?
The US can separate about 3,600 t/yr of NdPr oxide and, by the owners' statements, make about 7,610 t/yr of magnets at plants that run today, which is 29% and 20% of 2030 demand. If every plan were built, NdPr oxide would reach 10,840 t/yr and magnets 31,951 t/yr. Dysprosium oxide is the thinnest: none runs today and all stated plans reach 13% of demand.
Demand lines are Department of Energy forecasts (Feb 2022) as converted by the OSTI review, and dysprosium demand is derived from DOE's 1,900 t of embedded dysprosium (about 2,180 t as oxide). Metal and alloy has no published demand figure. The guide on why the US cannot process most of its own rare earths explains the gap.
How much NdPr oxide can the US separate?
The US can separate about 3,600 t/yr of NdPr oxide today, at Mountain Pass (about 2,600) and White Mesa Mill (850 to 1,000), which is 29% of 2030 demand. White Mesa's Phase 2 and the Round Top and Bear Lodge projects would add 7,240 t/yr for 10,840 t/yr, about 87%.
| Plant | Stated capacity (t/yr) | Status | When | Basis and notes | Source |
|---|---|---|---|---|---|
| Mountain Pass separations plantMP Materials, San Bernardino County, California | About 2,600NdPr oxide | Operating | Now | Reported outputMP made 2,599 t of NdPr oxide in 2025 and 917 t and 840 t in Q1 and Q2 2026. The OSTI review puts Mountain Pass and White Mesa together at about 3,600 t/yr, so Mountain Pass is that figure less the 1,000 t/yr at White Mesa. MP says its separations plant ran below design capacity in 2025 and expects full design capacity in Q1 2027; the 10-K gives no tonnage. | osti.gov sec.gov mpmaterials.com |
| White Mesa Mill, Phase 1 circuitEnergy Fuels, Blanding, Utah | 850 to 1,000NdPr oxide | Operating | Now | Stated capacityCounted at the top of the stated range. No annual output is published: the 2024 commissioning run turned 500 t of monazite into 38 t of NdPr product (company-reported). The circuit shares equipment with uranium milling, so rare earth and uranium runs alternate. | sec.gov sec.gov |
| White Mesa Mill, Phase 2 circuitEnergy Fuels, Blanding, Utah | About 5,000 moreNdPr oxide | Planned | By mid-2029 | DerivedThe January 2026 study and the 10-Q give over 6,000 t/yr for the whole mill after Phase 2. This row counts the increase over Phase 1 (6,000 less 1,000), so the mill is not counted twice. Phase 2 is not under construction: the company expects to finish it by mid-2029, subject to licensing, financing, feed and a final investment decision. Capital is about $410M. | sec.gov osti.gov |
| Round Top and Bear Lodge, combinedUSA Rare Earth, Rare Element Resources, Texas and Wyoming | About 2,240NdPr oxide | Planned | By 2030 (OSTI) | DerivedThe OSTI review reaches about 10,840 t/yr of NdPr oxide if all four projects run at 100% of design by 2030, against 8,600 t/yr for Mountain Pass and White Mesa at design. This row is the 2,240 t/yr difference. OSTI calls 100% of design within one to two years of start-up unlikely. USA Rare Earth targets first Round Top production in H2 2028 and Bear Lodge permitting is scheduled to finish in Mar 2028. | osti.gov |
| Marion advanced technology centerReElement Technologies, Marion, Indiana | Over 16,000 of products across four linesmixed oxides and products (not added) | Building | Q3 2026 to Q1 2027 | Company targetReleases give three different targets: up to 9,000 t/yr of oxide, over 10,000 t/yr (Jan 2026) and over 16,000 t/yr of aggregate product capacity. The figure counts all products, not NdPr, and no output is published. | sec.gov war.gov |
| Louisiana Strategic Metals ComplexUcore Rare Metals, Alexandria, Louisiana | About 9,600 of feed at full build; Machine A about 600TREO feed (not added) | Building | Machine A in H1 2027 | Study designCounts tonnes of rare earth oxide feed at full build (three lines of about 3,000 t/yr plus Machine A), not tonnes of NdPr. A demonstration plant began commissioning in Apr 2026. The sources checked name no commercial offtake customer. | newsfilecorp.com |
| Brook Mine refineryRamaco Resources, Sheridan County, Wyoming | About 3,400 of oxides (2025 assessment)mixed oxides and products (not added) | Planned | Initial production 2031 (conceptual study) | Study designFluor's 2025 preliminary assessment gave about 3,400 t/yr of oxides from about 5 Mt/yr of coal ore, with output targeted for the second half of 2028. The Jul 2026 conceptual study moved first production to 2031. The figure counts all oxides, not NdPr. | sec.gov |
| Elk Creek (rare earth products)NioCorp, Johnson County, Nebraska | About 650 on average over 40 yearsNdPr oxide (not added) | Planned | No date; construction target Dec 2026 | DerivedA long-run average: 25,923 t of NdPr oxide over a 40-year plan, divided by 40. It is not a capacity and no start date is published for rare earth output, so it is not added. | sec.gov |
| Port of Vinton separation facilityAclara Resources, Vinton, Louisiana | Not statedmixed oxides and products (not added) | Planned | Completion by end of 2027 | Company targetA $277M plant to separate heavy and light oxides from mixed carbonate mined in Brazil and Chile, with metals and alloys planned. Capacity is not stated in the sources checked. | opportunitylouisiana.gov |
| Gramercy bauxite residue projectElementUSA, Gramercy, Louisiana | 150 to 1,000mixed oxides and products (not added) | Planned | Demonstration facility construction from mid-2027; first production targeted Q3 2028 (company) | Company targetA plant of 150 to 1,000 t/yr of rare earth product from bauxite residue. DOE selected the project for award negotiation in Jun 2026; selection is not an award. | minesnewsroom.com |
| Seadrift plant (planned)Lynas USA, Seadrift, Texas | Not statedmixed oxides and products (not added) | Planned | No date; uncertain | Company targetLynas flagged considerable uncertainty about Seadrift in Aug 2025 and says it lacks the funds to complete it beyond the US money received. No decision has been published. | miningweekly.com |
The first four rows are added into the NdPr oxide total. The rest state feed, mixed products or an average, and are listed so that no plant is missing. OSTI rounds the all-built share to about 86%; the difference is rounding of the 12,500 t demand figure.
Can the US separate dysprosium and terbium?
Not at commercial scale yet. Energy Fuels is building circuits at White Mesa Mill for up to about 120 t/yr of dysprosium oxide and 20 t/yr of terbium oxide, due by the end of 2027. With Phase 2 the stated total is about 288 t/yr of dysprosium oxide and 80 t/yr of terbium oxide, 13% of the dysprosium demand DOE projects.
| Plant | Stated capacity (t/yr) | Status | When | Basis and notes | Source |
|---|---|---|---|---|---|
| White Mesa Mill, heavy rare earth expansionEnergy Fuels, Blanding, Utah | Up to about 120Dy oxide | Building | By end of 2027 | Company targetConstruction began on Jul 29, 2026 for about $104M. The dysprosium and terbium circuits are due by the end of 2027, with commissioning as early as Q4 2027; samarium, europium and gadolinium follow by the end of 2028. | sec.gov |
| White Mesa Mill, heavy rare earth expansionEnergy Fuels, Blanding, Utah | Up to about 20Tb oxide | Building | By end of 2027 | Company targetThe terbium circuit of the same expansion. White Mesa made its first kilogram of terbium oxide at 99.9% purity in Mar 2026. | sec.gov |
| White Mesa Mill, Phase 2 circuitEnergy Fuels, Blanding, Utah | About 168 moreDy oxide | Planned | By mid-2029 | DerivedThe 10-Q gives about 288 t/yr of dysprosium oxide for the Phase 1 and Phase 2 circuits together, so this row is 288 less the 120 above. The January 2026 study gave 240 t/yr for Phase 2 alone, which would make the total 360. The row uses the later and lower filing. | sec.gov |
| White Mesa Mill, Phase 2 circuitEnergy Fuels, Blanding, Utah | About 60 moreTb oxide | Planned | By mid-2029 | DerivedThe 10-Q gives about 80 t/yr of terbium oxide for both circuits together, so this row is 80 less the 20 above. The January 2026 study gave 66 t/yr for Phase 2 alone. | sec.gov |
| Mountain Pass heavy rare earth circuitMP Materials, San Bernardino County, California | Not statedDy oxide (not added) | Building | Output expected later in 2026 | Company targetThe first heavy rare earth separation circuit was mechanically complete in May 2026. MP expects dysprosium and terbium output later in 2026 and first samarium in 2028. No tonnage is given. | mpmaterials.com sec.gov |
| Elk Creek (rare earth products)NioCorp, Johnson County, Nebraska | About 66 on average (Tb about 17)Dy oxide (not added) | Planned | No date; construction target Dec 2026 | DerivedAverages over a 40-year plan (2,649 t of dysprosium oxide and 690 t of terbium oxide), not capacities. Not added. | sec.gov |
The Phase 2 rows are the increase over the Phase 1 expansion, to match the 10-Q's combined figures of about 288 t/yr of dysprosium oxide and 80 t/yr of terbium oxide. The January 2026 study gave 240 and 66 t/yr for Phase 2 alone. No terbium demand figure is published here, so terbium has no chart.
How much rare earth metal and alloy can the US make?
Plants state only 200 t/yr of metal capacity that is built, at Phoenix Tailings in Exeter, and 5,000 t/yr is under construction at USA Rare Earth's Blacksburg campus. MP Materials, USA Rare Earth at Stillwater and REalloys make metal or alloy but do not state a figure, so the total of 5,500 t/yr is a floor.
| Plant | Stated capacity (t/yr) | Status | When | Basis and notes | Source |
|---|---|---|---|---|---|
| Exeter metal refinery (built)Phoenix Tailings, Exeter, New Hampshire | 200metal and alloy | Operating | Now | Stated capacityCompany-reported: 200 t/yr built. Output and customers are not published. | fortune.com globenewswire.com |
| Blacksburg campus (metal and alloy)USA Rare Earth, Blacksburg, South Carolina | 5,000metal and alloy | Building | Commissioning 2028 | Company targetGroundbreaking was on Sept 9, 2026. The same campus is planned for 6,400 t/yr of magnets. | foxcarolina.com sec.gov |
| Exeter metal refinery (capacity not yet built)Phoenix Tailings, Exeter, New Hampshire | 300 moremetal and alloy | Planned | No date | DerivedThe company says the plant has 500 t/yr of capacity with 200 t/yr built; this row is the 300 t/yr difference. | fortune.com globenewswire.com |
| Independence (NdPr metal)MP Materials, Fort Worth, Texas | Not stated separatelymetal and alloy (not added) | Operating | Since Jan 2025 | Stated capacityMP has made NdPr metal at Independence since Jan 2025. Its metal and alloy capacity is not stated apart from the magnet capacity. | sec.gov |
| Stillwater (strip casting and metal)USA Rare Earth, Stillwater, Oklahoma | Not stated separatelymetal and alloy (not added) | Operating | Phase 1a since Mar 2026 | Stated capacityStrip casting and metal for USA Rare Earth's own magnets; part of the Stillwater magnet figure. | sec.gov |
| Euclid metallization facilityREalloys, Euclid, Ohio | Not publishedmetal and alloy (not added) | Operating | Now (company statement) | Stated capacityOperating status and capacity rest on company statements; neither output nor capacity is published. | sec.gov |
| Freedom FacilityPhoenix Tailings, Site not stated | Not statedmetal and alloy (not added) | Planned | Initial operations 2028 | Company targetA roughly $1B plant for separation and metallization of light and heavy rare earths, backed by a conditional $500M Office of Strategic Capital commitment (Jun 16, 2026). | globenewswire.com |
| Benson campus (metal)Vulcan Elements, Benson, North Carolina | Not stated separatelymetal and alloy (not added) | Planned | No date | Company targetThe planned campus is described as making magnets and metal; metal capacity is not stated apart from the 10,000 t/yr of magnets. | miningweekly.com |
Korean Metals Plant (Energy Fuels, South Korea) is outside the US and is not counted. Metal capacity is not stated apart from magnet capacity at Independence, Stillwater and Benson.
How many rare earth magnets can the US make?
Plants that run today state about 7,610 t/yr: Independence (about 3,000), Sumter (about 2,000), San Marcos (more than 2,000), Stillwater (600), Research Triangle Park (about 10). 10X and Blacksburg under construction add 13,400 t/yr, and Benson and HyProMag USA plan 10,941 t/yr more.
| Plant | Stated capacity (t/yr) | Status | When | Basis and notes | Source |
|---|---|---|---|---|---|
| Independence magnet plantMP Materials, Fort Worth, Texas | About 3,000NdFeB magnets | Operating | Projected; no date | Company targetThe 10-K says MP committed to expand the plant to a projected 3,000 t/yr of magnets. It began making magnets on commercial equipment in Dec 2025 and has delivered qualification samples to General Motors; commercial shipments are expected in Q4 2026. Output is not published. | sec.gov mpmaterials.com |
| Sumter NdFeB planteVAC Magnetics, Sumter, South Carolina | About 2,000NdFeB magnets | Operating | Now | Stated capacityVAC's release gives about 2,000 t/yr nameplate. Trade press reports the plant was built for 4,000 t/yr with a pathway to 12,000 t/yr; those figures are not counted. Commercial production began in fall 2025 and output is not published. | vacuumschmelze.com |
| San Marcos magnet plantNoveon Magnetics, San Marcos, Texas | More than 2,000 (target)NdFeB magnets | Operating | Target; no date | Company targetNoveon says it is scaling beyond 2,000 t/yr (Jan 2026). The row counts 2,000 as the floor of that target. Current output is not published. | noveon.co |
| Stillwater magnet plantUSA Rare Earth, Stillwater, Oklahoma | 600NdFeB magnets | Operating | Run rate expected by Q4 2026 | Company targetPhase 1a was commissioned in Mar 2026. 600 t/yr is the run rate the company expects by Q4 2026. USA Rare Earth states 10,000 t/yr for Stillwater and Blacksburg together, which leaves about 3,000 t/yr without its own figure; that part is not counted. | sec.gov |
| Research Triangle Park plantVulcan Elements, Research Triangle Park, North Carolina | About 10NdFeB magnets | Operating | Now (Oct 2025) | Stated capacityA 21,000 sq ft plant that opened on Mar 31, 2025. | manufacturingdive.com |
| 10X magnet campusMP Materials, Northlake, Texas | About 7,000NdFeB magnets | Building | Commissioning from 2028 | Company targetThe 10-K estimates 7,000 t/yr once completed and scaled, and 10,000 t/yr with Independence. The Pentagon buys the output for 10 years. | sec.gov mpmaterials.com |
| Blacksburg campusUSA Rare Earth, Blacksburg, South Carolina | 6,400NdFeB magnets | Building | Commissioning 2028 | Company targetGroundbreaking was on Sept 9, 2026. The campus is also planned for 5,000 t/yr of metal and alloy. | foxcarolina.com sec.gov |
| Benson magnet campusVulcan Elements, Benson, North Carolina | 10,000NdFeB magnets | Planned | Significant output by 2027 (company); full capacity undated | Company targetAbout $918M. The company says a significant share will be online by 2027; no date is published for the full 10,000 t/yr. A conditional $620M Office of Strategic Capital loan commitment to Vulcan was announced in the same month. | miningweekly.com |
| Dallas-Fort Worth recycling and manufacturing hubHyProMag USA, Denton County, Texas | About 941NdFeB magnets | Planned | Commissioning target mid-2027 | Study designSintered magnets made from recycled magnet material, from the updated design (the first feasibility study gave about 750 t/yr). Commissioning depends on final permitting and financing. | investegate.co.uk |
| Rochester technology centerArnold Magnetic Technologies, Rochester, New York | Not statedSmCo magnets (not added) | Operating | Now | Stated capacityArnold fabricates and tests samarium cobalt magnets in Rochester. Its US sintering capacity is not stated, and new SmCo capacity is in Thailand. | arnoldmagnetics.com |
Independence and Noveon figures are projections and targets, and the Stillwater figure is the run rate expected by Q4 2026. USA Rare Earth states 10,000 t/yr for Stillwater and Blacksburg together; the 3,000 t/yr not given its own figure is not counted. Arnold fabricates samarium cobalt magnets and states no US sintering capacity.
How much magnet recycling capacity does the US have?
Cyclic Materials' Mesa plant can take in up to 25,000 t/yr of end-of-life material, but that is material received, not rare earths recovered, and recovered output is not published. HyProMag USA plans about 941 t/yr of recycled magnets, which is counted in the magnet total.
| Plant | Stated capacity (t/yr) | Status | When | Basis and notes | Source |
|---|---|---|---|---|---|
| Mesa recycling plantCyclic Materials, Mesa, Arizona | Up to 25,000material taken in (not added) | Operating | Opened Sept 9, 2026 | Stated capacityCounts end-of-life material the plant can take in, not rare earth recovered. More than 7,000 t had arrived by opening, and recovered output is not published. | cyclicmaterials.earth |
More than 7,000 t of end-of-life material had arrived at Mesa by the opening on Sept 9, 2026. Noveon and MP Materials also use recycled feed; see the recycling guide.
How does this tracker count capacity?
It adds only stated capacity in one unit at a time, counts an increase and not a total twice, labels every figure as output, stated capacity, a company target, a study design or derived, and leaves foreign plants out. The six rules below apply to every row.
- Stated, not produced
Capacity is not output
Almost no company publishes magnet or metal output. MP Materials is the exception for NdPr oxide (2,599 t in 2025). Every other figure is capacity or a target.
- One unit at a time
Tonnes are not added across steps
Feed, mixed products, material taken in and magnets are different quantities. A row in another unit is listed and marked "not added" (16 of 36 rows).
- No double counting
Phases carry only the increase
Where an owner states a total that already includes an earlier phase, as at White Mesa Mill and the Exeter plant, the later row is the difference and its note shows the sum.
- Labeled basis
Targets say so
Each row is reported output, stated capacity, a company target, a study design or derived. A derived row says how in its note.
- US only
Foreign plants are left out
Energy Fuels' Korean Metals Plant (1,300 t/yr of alloy now, 3,600 t/yr targeted by end of 2026), Arnold's Thailand magnet plant and the Serra Verde mine in Brazil are outside the US and are not counted.
- Forecasts
Demand is a projection
The 2030 demand lines are Department of Energy forecasts under aggressive decarbonization scenarios, as cited by the OSTI review. They are not a measure of what buyers will take.
The five bases
- Reported output
- What the plant produced, used as its current level.
- Stated capacity
- The capacity the owner states for a plant that exists.
- Company target
- A capacity the owner has committed to or aims for. Not yet reached.
- Study design
- The design capacity in a feasibility study or engineering report.
- Derived
- Computed from stated figures; the note says how.
Where can I download the US rare earth capacity data?
The capacity data is free to download as CSV and JSON, one row per plant or project step, with the unit, status, basis, expected year, note and sources. It has 36 rows, verified through Oct 1, 2026, and may be used with a link back to this page.
| Column | What it holds |
|---|---|
id | A stable key for the row. |
step | The step of the chain: separated NdPr oxide, separated dysprosium and terbium oxide, metal and alloy, sintered NdFeB magnets or recycling. |
companies, facility, location | The owners (registry slugs), the plant or project and where it is. |
unit | What the figure counts. Only rows with the same unit are added. |
capacity_tonnes_per_year, capacity_as_written | The figure as a number (blank when none is stated) and as the source words it. |
counted_in_totals | true when the row is added into its step total. |
phase | Operating, under construction or commissioning, or planned. |
basis | Reported output, stated capacity, company target, study design or derived. |
year_expected, timing | The year the figure is expected to be in place (blank for operating plants and for plans with no date) and the timing as written. |
note, sources, verified | How the figure was read, the documents it comes from and the date it was checked. |
Where do I read more about US rare earth capacity?
The guides explain each step in the table, the timeline shows when plants are due, and the company lists show who runs each one.
- Why the US cannot process rare earthsThe gap between capacity and demand, and five obstacles
- TimelineWhen each plant and circuit is due, and which dates have moved
- Upcoming milestonesThe next dated steps, with how firm each is
- Mine to magnet supply chainWho does each step in the US
- Magnet manufacturersEvery US magnet maker and its plant
- Separation and refining companiesWho separates rare earths in the US
Questions about US rare earth capacity
How much NdPr oxide can the US make?
About 3,600 t/yr at Mountain Pass (about 2,600, MP Materials' 2025 output level) and White Mesa Mill (850 to 1,000), according to the OSTI review. Against about 12,500 t/yr of demand in 2030, that is 29%. Planned expansions would reach 10,840 t/yr, about 87% (the review rounds this to 86%).
How many rare earth magnets can the US make?
Operating plants state about 7,610 t/yr of NdFeB magnet capacity: Independence about 3,000, Sumter about 2,000, San Marcos more than 2,000, Stillwater 600, Research Triangle Park about 10. These are capacity statements and projections, not output, which almost no company publishes. Plants under construction and planned would bring the total to 31,951 t/yr.
Can the US separate dysprosium and terbium?
Not at commercial scale yet. Energy Fuels has made both at pilot scale and is building circuits at White Mesa Mill for up to about 120 t/yr of dysprosium oxide and 20 t/yr of terbium oxide, due by the end of 2027. MP Materials' heavy circuit at Mountain Pass is mechanically complete with no stated capacity.
How big is US rare earth demand in 2030?
The Department of Energy projected about 38,700 t of NdFeB magnet demand in 2030, which the OSTI review converts to about 12,500 t/yr of NdPr oxide, and about 1,900 t of embedded dysprosium, or about 2,180 t as oxide. These are forecasts under aggressive decarbonization scenarios, made in Feb 2022.
Is capacity the same as production?
No. Capacity is what a plant is rated or aiming to make. The OSTI review notes that projects rarely reach 100% of design within one to two years of start-up. Only MP Materials publishes NdPr output (2,599 t in 2025); the other rows are statements, and the page labels each one.
Which US magnet plant is the largest?
By stated capacity, Vulcan Elements' planned Benson campus (10,000 t/yr, no date for full capacity) and MP Materials' 10X campus under construction (7,000 t/yr, commissioning from 2028). Among plants running today, Independence (about 3,000 t/yr, a projection) and Sumter (about 2,000 t/yr nameplate) are the largest.
How current is the tracker?
The figures are verified through Oct 1, 2026. The directory re-verifies facts every quarter, scans for new announcements weekly and corrects errors within 48 hours of a report. The build checks that every plant on the map has a row here.
Sources for the capacity figures
Each profile lists its sources, a confidence grade and a verification date. Order of authority: official records (SEC filings, agency documents), company documents and technical reports, research papers, then independent press. Syndicated copies and reference works are labeled and are never the only source for a number. Company-reported figures are labeled as such, and non-binding deals are labeled "letter of intent" or "conditional".
- OSTI review, Minerals Engineering 249 (Kashyap et al., 2026) Research paper
- MP Materials 10-K for 2025 (SEC) Official record
- MP Materials Q2 2026 results Company document
- Energy Fuels 10-Q for Q2 2026 (SEC) Official record
- Energy Fuels 10-K for 2024 (SEC) Official record
- Ucore: engineering report for the Louisiana Strategic Metals Complex (May 2026) Company document
- American Resources press release on the Nov 2025 partnership (SEC exhibit 99.1) Official record
- Department of War: $25 million investment with ReElement Technologies (Jul 2026) Official record
- Opportunity Louisiana: Aclara invests $277M in Southwest Louisiana Official record
- Colorado School of Mines: $67M DOE award with ElementUSA Independent press
- Mining Weekly: Lynas flags uncertainty over Texas rare earths plant (Aug 2025) Independent press
- NioCorp 10-K for fiscal 2026 (SEC) Official record
- Ramaco 10-K for 2025 (SEC) Official record
- Fortune: Phoenix Tailings rare earth refinery in New Hampshire (Aug 2026) Independent press
- Phoenix Tailings: about $1B initiative to rebuild the US rare earth industrial base (GlobeNewswire, Jun 16, 2026) Company document
- Fox Carolina on the Blacksburg groundbreaking (Sept 2026) Independent press
- USA Rare Earth Q2 2026 results (SEC) Official record
- REalloys 8-K (SEC): merger and company description Official record
- Mining Weekly on the Benson plant (Nov 2025) Independent press
- VAC: opening ceremony for Sumter (Jul 2026) Company document
- Noveon: Series C announcement Company document
- Manufacturing Dive on the Benson plant Independent press
- Mkango: HyProMag USA provides a positive update to valuation Company document
- Arnold: who we are Company document
- Cyclic Materials: opens the first commercial-scale automated rare earth magnet recovery facility in the US (Sept 2026) Company document
- Department of Energy: Rare Earth Permanent Magnets, Supply Chain Deep Dive Assessment (Feb 2022) Official record
Tags show what kind of source each link is. How sources are ranked