# Zhongke Sanhuan (Beijing)
Beijing Zhong Ke San Huan; vertically integrated sintered NdFeB (~12k+ t/yr). General export license (Dec 2025). H1-2026: domestic motor-magnet partnerships; no overseas plant.
Source: REEx
Insights · Updated October 2026
Rare-earth magnet supply chain atlas
Who does what, where — from ore to finished magnet. Seven process steps, 143 organizations, 27 countries, every claim sourced. Sintered rare-earth magnets only: NdFeB and SmCo.
NdFeB
SmCo
Both
JV / partner / subsidiary
Under construction (▸ target year)
Under consideration (◇ year — pre-FID, permitting or raising money)
Recycling: recovers third-party material
Recycling: own production scrap only
Unverified
Show
Work with Peculiar Materials
I keep this map current because my own work depends on it. If you are evaluating a magnet or rare-earth company, qualifying a supplier, or trying to find out whether a plan survives contact with a factory floor, that is the work I do. The weekly digest is where a change like one of these shows up first.
Producer directory
Every organization on the chart above, in full: what each one does in-house, where it sits on the eight steps, whether it is running today or still being built, and the public source behind each claim. 143 organizations across 27 countries. Every entry has its own link.
China 18 organizations
# Konit
Ningbo · San Huan co.
Separate San Huan company in Ningbo. Sintered NdFeB midstream→assembly.
# Ningbo Yunsheng
Sintered NdFeB (Ningbo + Baotou) ~26k t/yr; also bonded. General export license (Dec 2025). Sep 2026: DELAYED the remaining 10,000 t of its 15,000 t Baotou expansion (Jul 2026, policy/macro uncertainty); pursuing HK listing. No Thailand/Vietnam plant found (SE-Asia trade hits are bonded exports).
Source: REEx Jul-26
# HPMG (Hangzhou Perm. Magnet)
Full sintered NdFeB + SmCo + AlNiCo + assemblies; GBD / low-HRE grades.
Source: chinahpmg
# Chengdu Galaxy Magnets
Makes SmCo magnets and assemblies (Shenzhen ChiNext: 300127), alongside bonded and hot-pressed NdFeB magnets from its own fast-quench powder. The atlas covers sintered magnets only, so the row is drawn for its SmCo magnet production.
Source: galaxymagnets
# Ningbo Ninggang (NGYC)
SmCo specialist (1:5 & 2:17) since 1987 (~1,500 t/yr); also NdFeB/AlNiCo/ferrite.
Source: ngyc
# XMag (X-Magnet) — Fenghua, Ningbo
Private · Fenghua, Ningbo + Bac Ninh, Vietnam · magnet machining
XMag Co. Ltd (X-Magnet): office in Taipei, Taiwan; factories in Fenghua District, Ningbo, Zhejiang, and Bac Ninh, Vietnam (XMag Technology Co., Ltd, registered Oct 2023). Its website lists machining equipment (400+ slicing, 180 automatic drilling, 30 wire-cutting and 300 custom grinding machines), more than 200 direct employees, a 10,000-square-meter main campus and a 2012 output value of CNY 130 million. No tonnage, pressing or sintering line, or magnet material is stated. At its current site since 2008; X-Magnet brand adopted 2013. Applications named: wind turbine generators, e-bike and e-motorcycle motors, sewing machines, consumer-electronics latches and Hall sensor modules. Privately held; no ownership, customers or 2025-2026 figures published. Not Hangzhou X-Mag Inc.
Source: XMag website Vietnam tax record
# Earth-Panda
Hefei, Anhui · 688077
Anhui Earth-Panda Advanced Magnetic Material (STAR 688077; Lujiang/Hefei). Sintered NdFeB (main) + SmCo + bonded; machining/surface/recycling + assemblies. National key lab for RE permanent magnets. Recycling: H1 2025 R&D projects recover rare earth from its own grain-boundary-diffusion machining sludge and from end-of-life sintered NdFeB and SmCo product. Disclosed as R&D, not as operating capacity; no tonnage published.
Source: earth-panda FT
# JL MAG
World's largest high-performance NdFeB (Ganzhou/Baotou/Ningbo; 5 plants). Sep 2026: ~40,000 t/yr at end-2025; Baotou Phase 3 (+20,000 t) under construction → ~60,000 t/yr by end-2027. MEXICO project (2023, US$100M; re-scoped Oct 2023 to 1M magnetic assemblies/yr) is ABSENT from the FY2025 annual report — treat as STALLED. ~63% feed from national RE groups; ~30% recycled. No Vietnam plant. Recycling: 5,000 t/yr approved capacity at 51%-owned Yinhai New Materials (Inner Mongolia) for comprehensive utilization of rare-earth product waste; 3,681 t of rare-earth raw material actually recovered in 2025. Feed is its own magnetic sludge and offcuts plus purchased end-of-life magnet steel.
Source: JLMAG AR25 REEx Jul-26 Mexico
# Yantai Zhenghai Magnetic
300224 · Yantai + Nantong
Sintered NdFeB only. ~30,000 t/yr blank capacity (Yantai 18k + Nantong/Rugao 12k; ~85% utilization 2025). GBD heavy-RE-lean grades; >8M EV motor sets/yr; exports ~40% of revenue (+43% 2025). Earlier 36k-by-2026 target softened. NO overseas plant found (a 'Mexico' lead is unsubstantiated).
Source: REEx Jul-26 STCN
# Baotou Tianhe Magnetics
Baotou + Tianjin · sintered SmCo too
Sintered NdFeB (~12,500 t/yr capacity; ~6,200 t output) AND sintered SmCo (~500 t/yr) — one of few Chinese majors with SmCo at scale. Strip-casts in-house and builds its own magnet-making equipment. CNY 900M Inner Mongolia R&D/production base: phase-1 construction from Mar 2026 (~18 months). Offices JP/US/DE.
# Golden Dragon Rare Earth
Xiamen Tungsten · Longyan + Baotou
Wholly owned by Xiamen Tungsten (600549). Genuinely vertical: RE SEPARATION (Fujian/Jiangxi) → metal → alloy → sintered NdFeB blanks → machining. Baotou phase I 5,000 t/yr commissioned Jun 2025 (RMB 462M) toward 20,000 t/yr; further output expansion reported Jul 2026. State-linked national champion. Recycling: subsidiary Yuantong New Materials is engaged in rare-earth secondary-resource recycling (H1 2025 NEEQ report). No capacity or feedstock breakdown published.
Source: REEx Metalnomist
# AT&M (Advanced Technology & Materials)
000969 · Beijing + Antai Beifang (Baotou)
Central-SOE-backed (CISRI orbit). ~12,000 t/yr sintered NdFeB capacity (Feb 2026), incl. Antai Beifang 5,000 t/yr high-end JV with Northern Rare Earth (ramping). >10,000 t shipped 2024. SmCo NOT confirmed in its own disclosures — shown NdFeB only.
# Zhejiang Innuovo Magnetics
000795 · Hengdian Group
Hengdian (Dongyang) — sintered NdFeB ~13,000 t/yr blanks (~80% utilization) +5,000 t expansion from H2-2025 → ~18,000 t/yr; also motors downstream. Bonded NdFeB/ferrite not shown. Has an active LinkedIn company page.
# China Northern Rare Earth
600111 · Bayan Obo / Baotou — upstream anchor
World's largest LREE miner (Bayan Obo) + largest single separator + ~100,000 t/yr magnetic-alloy (strip-cast) capacity (Jun 2025) — the biggest alloying node on earth. Magnet making only via JVs (not drawn): Antai Beifang (w/ AT&M, 5,000 t), Northern Zhaobao (w/ Ningbo Zhaobao, 3,000 t, first batch Oct 2025), Northern Funeng (10,000 t RE metals/FeDy/FeHo, announced May 2026, schedule undisclosed). Recycling: 4,960 t/yr scrap-recovery capacity at subsidiary Xinfeng Xinli, plus 4,000 t REO/yr at Jinmeng under construction. Since 2022 its supply agreements require downstream magnet makers using its PrNd metal to return their production scrap for recovery — a contractual claim on other companies' swarf.
Source: Sina Metalnomist REEx
# China Rare Earth Group
heavy-RE duopoly half · Ganzhou
State group formed 2021: ionic-clay mining (Jiangxi/Guangdong — the Dy/Tb half of China's duopoly), separation, metallization, magnet-alloy production, recycling; GRINM/Grirem R&D arm. Subsidiaries: China Southern RE, Guangdong RE, listed 000831 & 600259. Capacities not disclosed by site. Dominant in Dy/Tb (no non-Chinese supply of scale) and Sm. Recycling: 6,600 t/yr REO from NdFeB scrap at subsidiary Ganzhou Youli, launched Dec 2023 as a 3.7x expansion and billed as China's largest single-site NdFeB scrap recycler. No commissioning announcement has appeared; Youli sits under the unlisted group, so no disclosure obligation forces one.
# Jiangsu Huahong Technology
002645 · Ganzhou + Ji'an recycling · Ningbo magnets · 12,000 t/yr recycled REO
Shenzhen-listed (002645) group that recovers separated rare earth oxides (Pr, Nd, Dy, Tb, Gd, Ho and others) from NdFeB production scrap and end-of-life magnets at one pre-treatment base and three smelting-separation bases in Ganzhou and Ji'an, Jiangxi (subsidiaries Ji'an Xintai, Jishui Jincheng, Jiangxi Wanhong). Stated oxide capacity is more than 12,000 t/yr; actual 2025 production was 8,794 t of oxides (+71% YoY). Subsidiary Zhonghang makes sintered NdFeB in Ningbo and Liangshan: stated capacity 20,000 t/yr, 2025 production 15,792 t. A 10,000 t/yr magnet plant in Baotou was slated for trial production in Q2 2026 and was still under construction in August 2026. FY2025 revenue was RMB 7.84bn, of which rare earth recycling RMB 4.08bn (52%) and magnets RMB 1.91bn (24%); H1 2026 revenue was RMB 5.60bn, with rare earth recycling at 65%.
Source: H1 2026 report FY2025 annual report
# Zhongxi Tianma — Liangshan, Shandong
Private · Liangshan, Shandong · 36,000 t/yr scrap feed (stated)
Zhongxi Tianma New Materials Technology (中稀天马), Liangshan County, Shandong, registered August 2010. Recovers rare earths from NdFeB magnet scrap and sells PrNd, Gd, Tb, Dy and Ho oxides and rare earth metals. Company-stated capacity: 36,000 t/yr of rare earth waste processed (a 2022 article gave 30,000 t/yr); a December 2024 Jining report gave 10,000 t/yr annual capacity, product unstated. Privately held: chairman Lin Ping held 41.07% (Jan 2024); Sanchuan Wisdom (SZSE: 300066) acquired 20% in 2019. Quotation on the NEEQ (code 831671) ended August 20, 2020; IPO tutoring was terminated in September 2025. Last published accounts: 2019 revenue CNY 530 million. A 4,000 t/yr scrap pretreatment center in Baotou is listed as planned for 2026. No public source links the company to China Rare Earth Group; it used the Zhongxi Tianma name by April 2019, before the group's December 2021 formation.
Source: Times Finance via eastmoney (Aug 2026) 10jqka (Aug 2026)
Japan 8 organizations
# Shin-Etsu Chemical
Integrated: RE-oxide separation/refining → melt/cast/sinter → machining → magnetic-circuit assembly. NdFeB + SmCo. Footprints Japan/China/Vietnam (SMMV, under Vietnam). Sep 2026: Daikin partnership to recycle AC-compressor magnets from 2027. New Fukui refinery = separate row. Recycling: collects magnets generated in its own NdFeB manufacturing process and used magnets returned from customers' processes. Reported as industrial-scale hydrometallurgy running since 2008 in Japan and 2013 in Vietnam; volumes undisclosed. Also the magnet manufacturer in the Daikin/Hitachi/Tokyo Eco Recycle air-conditioner compressor consortium. METI's own certification record (separate critical-minerals and permanent-magnet tracks) confirms the Fukui JPY 17.5bn subsidy (certified 19 May 2026, three weeks before the company's own 11 Jun 2026 press acknowledgment) and two SMMV/Thailand/Philippines permanent-magnet certifications (Apr 2024, ~¥13bn; Apr 2025, amount undisclosed).
Source: Shin-Etsu note METI critical minerals METI permanent magnets
# Shin-Etsu — Fukui refinery
new domestic RE refining
Announced 11 Jun 2026: ~¥35bn (~US$218M) rare-earth refinery/smelter in Fukui — Shin-Etsu's first Japanese RE plant in 18 years, explicitly to cut China dependence after the Apr 2025 export controls. Capacity and start date NOT disclosed; ▸2028 is a provisional placeholder.
Source: Japan Times Fastmarkets
# Proterial (ex-Hitachi Metals)
Invented sintered NdFeB (1982); NEOREC (ferrite excluded). Sep 2026: Proterial (India) Ltd is a bidder in India's ₹7,280 cr REPM scheme (bids opened 13 Aug 2026). Historic China Grove NC (US) site — no evidence of an active sintered line in 2025-26.
Source: note India bids
# Daido Steel / Daido Electronics
NdFeB: hot-deformed (NEOQUENCH-DR, HRE-free) + isotropic bonded; own rapid-quench powder. SmFeN NOT colored as SmCo.
# TDK
buys alloy (e.g. Santoku)
NdFeB + ferrite (ferrite not shown). Starts from PURCHASED alloy → jet-mill, press, sinter, machine, magnetize. Sep 2026: acquired Showa Denko/Resonac's NdFeB magnet-ALLOY R&D business — internalizing alloy know-how, but still shown as buying alloy.
Source: MagneticsMag
# Santoku
alloy supplier — no magnets
Materials supplier (NOT a magnet maker) — direct competitor to Less Common Metals (USARE UK sub). Strip-cast NdFeB alloy + SmCo alloy/powder + RE metals + recycling (Kobe/Miki/Akashi/Tsuruga). Sep 2026: the separation build is a JOGMEC ~¥0.8bn FY2026 grant for domestic refining equipment (mixed RE carbonate → NdPr oxide); site/route/capacity undisclosed. Now traced to METI's own record: 2026重要鉱物第2号-1, ~¥800m, certified 29 Jul 2026. Santoku also holds an earlier, separate METI permanent-magnet-track certification (2024永久磁石第3号-1, 24 Jul 2024, amount undisclosed) — moved ▸2026→▸2027, low visibility. Recycling: a standing Recycling Division recovers neodymium magnet grinding and polishing powder from users or from the open market and converts it back to rare-earth alloy feedstock — third-party feed, not its own scrap. No tonnage published.
Source: Santoku CMF METI critical minerals METI permanent magnets
# Daikin + Shin-Etsu + Hitachi + Tokyo Eco Recycle
consortium · commercial AC compressor magnets · from 2027
Announced April 14, 2026 in a joint release: a collaboration to recover rare earth magnets from compressors of Daikin commercial air conditioners replaced during repairs and overhauls in Japan, a stream with no legal collection framework (household units are covered by the 2001 Home Appliance Recycling Act). Daikin collects compressors; Tokyo Eco Recycle (Koto-ku, Tokyo; Hitachi holds a stake) disassembles, demagnetizes and extracts magnets, with Hitachi on AI and robotic disassembly; Shin-Etsu Chemical makes new magnets from the recovered material. Automation equipment is in development during 2026; full-scale operation is targeted from 2027. The companies give no volume, capacity or investment figure, and no binding terms are disclosed. The participant is Hitachi, Ltd., not Proterial.
# DOWA Eco-System
TSE 5714 sub · appliance and EV motor magnet recovery demos
Wholly owned subsidiary of DOWA Holdings (TSE: 5714). Since November 2025 it has run a demonstration test at its Akita group company Eco-Recycle in Odate, recovering neodymium magnets from motors of dismantled home appliances by demagnetizing the motors and removing the magnets, for reuse as raw material for new magnets (release Nov 20, 2025). On Sept 10, 2026, DOWA Holdings said a DOWA Eco-System proposal to demonstrate magnet recycling from end-of-life EV drive motors was selected under the Ministry of the Environment's FY2026 public call; no site, schedule, budget or partner was named. No capacity, tonnage or commercial plant has been announced.
Source: DOWA HD 10-Sep-26 DOWA Eco 20-Nov-25
Korea 5 organizations
# Star Group (SGI) + POSCO Int'l
ALLIANCE (one row). Star Group Industrial / Sung Lim (Daegu, 1994) = Korea's leading sintered-NdFeB producer (~1,000 t/yr Daegu); POSCO Int'l = trading/sales + drive-motor cores. Sep 2026: preparing KOSDAQ listing (NH lead underwriter); low-RE magnets co-developed with Hyundai. US plant STALLED (see United States); Vietnam plant SGI Vina operating.
Source: POSCO G-Enews Jul-26
# LS Cable & System — Korea magnet plant
Korea Motion Works · Gumi · KRW 150bn · no capacity yet
Korean magnet plan, now held by Korea Motion Works, a company LS Cable & System set up in Sep 2026 to make and sell rare-earth permanent magnets (100% LS Cable; not LS Eco Advanced Materials). On 30 Sep 2026 LS Cable disclosed a KRW 150bn equity injection (150M shares), paid in installments Oct-Dec 2026. Press reports Gumi, North Gyeongsang as the site (Khan: leading candidate). Target uses: EV, robot, wind and defense motors; reports cite Nd- and Sm-based magnets. No capacity, construction start or production date published (as of 16 Oct 2026). Planned feed: Sm metal (later NdPr/Dy/Tb) from LS Eco Energy's Vietnam plant, made from Lynas oxide. The Feb 2024 plan for a 1,000 t/yr NdFeB magnet JV between LS Eco Advanced Materials and VAC (production by 2027) has had no public update. US magnets: separate LS Cable Chesapeake row.
Source: SEDaily Sep-26 Khan Sep-26
# JS Link
Yesan · 1,000 t/yr · Lynas feed
Sintered NdFeB magnet plant in Yesan, South Chungcheong, South Korea: 1,000 t/yr magnet nameplate (company figure), with in-house strip-cast alloy flake, sintering, heat treatment, grain boundary diffusion and post-processing. A ~1.8 t trial batch (42SH) was reported in Oct 2025; KIMS verified N52 (Mar 2026) and 45UH (Jul 2026) samples. First mass-production order (4 Sep 2026): 360 kg of 45SH for a North American robot-parts maker, about US$120,000, deliveries from Dec 2026. An alloy-flake line expansion was due Oct 2026. Lynas is exclusive rare earth supplier to Jan 2038 and bought ~4.58% of JS Link for A$50M (closed 29 Jul 2026). Non-binding LOI with REalloys (Jul 2026) on a North American platform. As of 2Q 2026, ~99% of company revenue was genome-analysis services.
Source: Lynas ASX MT/The Bell
# Energy Fuels — Korean Metals Plant US-owned
ex-ASM · Ochang
Australian Strategic Materials' Ochang plant — acquisition by Energy Fuels COMPLETED 28 Aug 2026. Commercial NdPr metallization + NdFeB strip-cast alloy: 1,300 t/yr alloy today → 3,600 t/yr expansion 'as early as end-2026'; Dy/Tb metallization in development. Largest ex-China/Japan metal-alloy node in Asia. EF plans a replica 'American Metals Plant' (no site/date; InvestorNews reports an initial 2,000 t/yr of alloy, expandable to 4,000 t/yr, though EF's own release gives no figure). Named orders (ASM, 6 Jun 2025): Vacuumschmelze 7.2 t NdFeB alloy (binding purchase orders, delivered from Jul 2025), Noveon under an existing 100 t agreement, Magnequench 19 t NdPr metal.
Source: EF 28-Aug-26 ASM KMP ASM: Sales momentum builds at KMP
# Evolution Metals & Technologies — Pohang
Pohang · 660 t/yr disclosed · going-concern doubt
Nasdaq: EMAT, listed through a SPAC merger that closed 5 Jan 2026. The 10-K discloses 660 t/yr of magnet capacity (400 t/yr from 2007 plus 260 t/yr from 2024) and 144 t/yr NdPr metal / 216 t/yr NdPr alloy (2021). Press releases claim about 1,000 t/yr today and 10,000 t/yr by end-2026; neither figure is in a filing. No tonnage produced or shipped has been disclosed. H1 2026 revenue $3.5M. Q2 2026 10-Q: $5.3M cash, $78.8M working capital deficit, substantial doubt about continuing as a going concern not alleviated. A claimed 130 MW to 750 MW power expansion has no utility or government confirmation. The issuer pays for ongoing stock promotion (disclosed on the promotional articles themselves).
India 6 organizations
# IREL (India Rare Earths)
Odisha · Kerala · Vizag · state-owned · 11,200 t/yr RE chloride
India's 100% government-owned rare earth producer under the Department of Atomic Energy; monazite processing is reserved to it under the Atomic Energy Act. Beach-sand mining and mineral separation at Chavara (Kerala), Manavalakurichi (Tamil Nadu) and OSCOM (Odisha). OSCOM extraction plant: 11,200 t/yr mixed RE chloride installed (2015-16). Aluva refinery makes >99% oxides incl. NdPr and Sm. FY2024-25: RE chloride output down ~11%; Aluva curtailed for lack of domestic buyers. NdPr oxide capacity reported at 400-500 t/yr (SMM, Aug 2026). Vizag plant (inaugurated May 2023; ~3 t/yr per Parliament, Apr 2026) has started SmCo magnet production. 2012 Toyotsu export deal reported suspended June 2025. Assured NdPr oxide for the 3 lowest bidders in the Rs 7,280 crore, 6,000 t/yr magnet scheme (20 bids, Aug 2026).
# Midwest Advanced Materials (MAM)
Hyderabad · 500 t/yr oxide-to-magnet · commissioning
Hyderabad, Telangana. Integrated sintered NdFeB plant that starts from rare earth oxide, with molten-salt electrolysis metallization transferred from NFTDC (Hyderabad). Initial nameplate 500 t/yr of magnets from about 150-170 t/yr of oxide, which the company said would come from IREL (June 2025). India's Technology Development Board backs it with INR 125 crore toward an INR 250.86 crore project (2024). A December 2025 first-product target (June 2025) was not met; on Sep 28, 2026 the group, reported as Midwest Energy, said the plant was being commissioned, with commercial production targeted by mid-October 2026 and expansion to 5,000 t/yr by December 2027. Samples are with OEMs; no offtake announced. A Midwest Energy-MAM joint venture bid in India's REPM scheme (Aug 2026); no award yet.
Source: ET via Whalesbook Sep-26 BS Jun-25
# India REPM scheme (6,000 t/yr tender)
₹7,280 crore · 6,000 t/yr · 20 bids, no awards
Government of India program, not a company. Cabinet-approved Nov 26, 2025: INR 7,280 crore (INR 6,450 crore sales-linked incentives over 5 years, INR 750 crore capital subsidy) for 6,000 t/yr of integrated sintered NdFeB capacity, 600-1,200 t/yr per beneficiary, up to five. The RFP (Mar 20, 2026) requires all steps from NdPr oxide to sintered magnet in-house, ranks bids by lowest incentive (cap INR 2,150/kg) and requires 50% of capacity within 3 years of award. IREL oxide (400 t/yr capacity) is indicative and non-binding. After three extensions, 20 technical bids were opened Aug 13, 2026, including Coal India, L&T, Neo Performance (Singapore), Proterial India, Attero, Lohum, ReNew, 20 Microns and a Midwest JV. No award as of Oct 19, 2026.
Source: MHI RFP Mar-26 PSU Watch Aug-26
# Attero Recycling
Noida · e-waste magnet recycling · 30,000 t/yr target
Noida, Uttar Pradesh; private e-waste and battery recycler founded 2008, main plant at Roorkee, Uttarakhand. Recovers Nd, Pr and Dy (Ce and Gd also present) from magnets in hard disk drives, laptops and earphones. On Jun 19, 2025 it announced INR 100 crore to scale rare earth recycling from 300 t to 30,000 t a year (1 to 100 t/day) within 12-24 months; whether this measures feed or recovered rare earths was not stated, and in Sep 2026 the CEO cited about 100 t/yr of rare earth output. Recovery (98%) and purity (99.9%) are company figures; product form is not published. Feb 2026: plan for about INR 7,000 crore over five years in refining, sites not chosen. One of 20 bidders in India's REPM magnet scheme (bids opened Aug 13, 2026); no award as of Oct 19, 2026.
Source: BS Jun-25 Swarajya Feb-26
# Lohum unverified
Greater Noida · oxide-to-magnet plan · ECMS-approved
Noida, Uttar Pradesh; private lithium-ion battery recycler and critical minerals company founded 2017. Plans an integrated plant at Site 5, Greater Noida, to make sintered NdFeB magnets from rare earth oxide. Announced Nov 10, 2025 as 2,000 t and INR 500 crore, targeting up to 20% of Indian magnet demand by 2027; approved Mar 30, 2026 under MeitY's Electronics Components Manufacturing Scheme (25% capex subsidy expected). A May 12, 2026 report gave 5,000 t/yr and INR 700 crore and called it a pilot facility. Jul 28, 2026: MoU with CSIR-IMMT to develop NdPr metallization. Plans Dy and Tb recovery from e-waste; recycled share of feed not specified. No NdPr oxide supply deal announced. Lohum Magnets & Energy Solutions bid in India's REPM scheme (Aug 2026); no production or award reported as of Oct 19, 2026.
Source: Business Today Mar-26 KNN May-26
# N.A.N. MagneTech — Naidupeta, Andhra Pradesh
integrated oxide-to-magnet plan · 1,200 t/yr Phase I · REPM bidder
Private; incorporated 13 Apr 2026 under N.A.N. GreenMet, founded and chaired by Navin Agarwal, Executive Vice Chairman of Vedanta, who is also listed as a director of N.A.N. MagneTech (Vedanta itself is not a party in any source). Announced 23 Jul 2026: Rs 1,250 crore Phase I at Naidupeta Industrial Park, 1,200 t/yr sintered NdFeB rising to 10,000 t/yr, covering oxide to metal, alloy and magnets plus recycling of production scrap; commercial production Q1 2028 (state order: June 2028). Andhra Pradesh granted 25 acres (Jun 2026). One of 20 bidders for India's REPM incentive scheme (five awards of up to 1,200 t/yr each; none announced as of 29 Sep 2026), so not yet part of the 'REPM scheme awardees' row. Technology partner (Japanese), oxide supplier (Australian) and customers unnamed.
Source: N.A.N. MagneTech to Build Rs 1,250 Crore Andhra Pradesh Approves 2 Rare Earth Mag Vedanta leadership (Navin Agarwal's directorships)
United States 42 organizations
# MP Materials
Mountain Pass (mine+sep) → Independence, Fort Worth TX: molten-salt metal, strip-cast alloy, sinter, machine, magnetize. Sep 2026: Independence magnet production began Dec 2025 (~1,000 t/yr; Q2-2026 still customer-qualification deliveries); Q2-2026 sold 1,006 t NdPr. DoW: $400M pref. equity + 100% offtake on 10X + loans (Jul 2025). Saudi Ma'aden refinery JV = binding term sheet only (Saudi section). 10X = separate row. Recycling: a commercial-scale line at Mountain Pass under construction to process post-industrial and end-of-life magnets and magnet scrap, alongside short-loop and long-loop circuits in Texas and California taking scrap from MP's own magnet production. No recycling throughput published — the 10,000 t/yr figure is magnet capacity.
# MP Materials — 10X (Northlake, TX)
10,000 t/yr campus
Site selected 26 Feb 2026: Northlake TX (120 acres, <10 mi from Independence). >$1.25B + ~$200M TX incentives; ~10,000 t/yr NdFeB incl. GBD; 1,500+ jobs; closed-loop recycling. DoW 100% offtake. Commissioning begins 2028; CEO (Aug 2026): 'construction accelerated'. Largest announced Western magnet plant. Recycling: scrap from Texas magnet production feeds MP's short-loop and long-loop recycling circuits in Texas and California. No third-party feedstock attached to this campus in the public release.
# USA Rare Earth
Stillwater OK (u/c) + Round Top TX
Stillwater OK sintered NdFeB — drawn UNDER CONSTRUCTION: Phase 1a line commissioned 26 Mar 2026 but not at commercial output; 600 t/yr run-rate targeted Q4-2026; local press disputes the ramp. Round Top TX mine ▸late-2028 (pulled forward 2 yrs; DFS Q4-2026; 100% owned after TMRC deal Aug 2026). Dashed = subsidiaries/partners: Less Common Metals (UK, own row), Wheat Ridge CO demo (Dy/NdPr oxide from scrap), Carester FR (13.6% stake Jul 2026). Sep 2026: SERRA VERDE (Brazil mine) combination COMPLETED 4 Sep 2026 — see Brazil. Blacksburg SC plant = separate row. Q2-2026 revenue miss = funding-risk watch item. Recycling: the Wheat Ridge, Colorado hydromet demonstration line, commissioned June 2026, has produced commercial-grade Dy and NdPr oxide from NdFeB swarf generated at its own Stillwater plant. The company projects swarf could support up to 30% of future feedstock needs — a projection, not capacity. End-of-life magnets named only as a future option.
Source: USARE Q2-26 Serra Verde
# USA Rare Earth — Blacksburg, SC
6,400 t magnets + 5,000 t alloy
Site selected 2 Jun 2026; GROUNDBREAKING 9 Sep 2026. Bailey Industrial Park, Cherokee County SC: 124 acres / 800,000 sq ft, ~$1.2B, ~490 jobs. 6,400 t/yr sintered NdFeB + 5,000 t/yr strip-cast metal & alloy; commissioning 2028. With Stillwater expansion → 10,000 t/yr USARE magnet capacity.
Source: GNW 9-Sep-26
# Noveon Magnetics
San Marcos TX; only long-running US sintered NdFeB maker (~2,000 t/yr); recycled + primary feed (Lynas, Solvay). Customers GM, ABB Motion, Nidec (>1,000 t binding). $215M Series C (Jan 2026, OneIM) to expand San Marcos 'beyond 2,000 t' — on-site, no dated capacity target (10,000 t ambition undated). Korea: 2,000 t/yr plant MOU w/ Kangwon Energy (Nov 2025, MOU only); LG closed-loop recycling. Recycling: feedstock is recovered end-of-life magnetic material from motors, energy systems, MRI, data storage and hybrid/EV drivetrains. The recycled fraction is not published — Noveon states it can manufacture from virgin material, recycled content or a combination.
# Vulcan Elements + ReElement
LAB SCALE today (~10 t/yr Durham)
ALLIANCE (one row). Vulcan = sintered NdFeB at LAB / pilot scale only (Durham NC, ~10 t/yr; not a commercial producer today) — commercial capacity is the Benson NC plant (▸2028, separate row). ReElement (dashed) = Marion IN separation/refining + recycling — minor-source air permit for all lines (19 Aug 2026), $25M DoW award (Jul 2026), first Ge column Aug 2026. Benson NC 10,000 t plant = separate row. NB: ReElement also in the POSCO separation JV (below). Recycling: ReElement refines rare earth from secondary feedstock it did not produce — end-of-life magnets, hard drives, EV motors, wind turbines, industrial scrap, plus e-scrap aggregated by ERI — and takes back Vulcan's own magnet manufacturing scrap. Marion Phase 1 announces a line of up to 7,000 t/yr separated oxide from recycled feedstocks.
# Vulcan Elements — Benson, NC
10,000 t/yr · $1.4B US-gov't package
Johnston County NC selected 18 Nov 2025: ~$918M–1B, 1.1M sq ft, 10,000 t/yr, ~1,000 jobs. Funding: $620M DoD Office of Strategic Capital loan + $50M Commerce equity + $120M NC incentives + ~$550M private (Mar 2026). Ucore feedstock deal (Mar 2026). NO groundbreaking announced as of Aug 2026 (site page still future-tense); ~1,000× scale-up from today = largest execution risk on the map. ▸2028 first magnets.
Source: PRN Nov-25 Benson page Axios
# Energy Fuels / eVAC
Energy Fuels: monazite (FL/GA) → White Mesa UT sep (Phase 2 → 6,000 t NdPr by mid-2029; Dy/Tb circuits under construction since 29 Jul 2026, due end-2027; US heavy-REO qualified for magnets Dec 2025). eVAC Sumter SC (dashed) = block→assembly NdFeB+SmCo, OPERATING since Dec 2025 (2,000 t/yr nameplate, ~1,500-1,600 t practical; expandable to 4,000). Offtake terminated: the customer gave notice on 1 Jul 2026 and e-VAC has been in arbitration since 7 Jul (Energy Fuels 8-K, 13 Oct 2026); GM's Q3 2026 10-Q says it terminated a strategic sourcing agreement with an unnamed EV traction-motor component supplier, and Reuters (23 Oct 2026, citing a person with knowledge) reports GM canceled its Sumter supply deal. Sep 2026: VAC acquisition ($1.9B, agreed 23 Jun 2026) NOT closed — 'end-2026' (Q2 release) vs 'early 2027' (deal release). ASM (Korean Metals Plant) acquisition COMPLETED 28 Aug 2026 → EF now a mine-to-metal-alloy group; planned US 'American Metals Plant' (no site/date; InvestorNews reports an initial 2,000 t/yr of alloy, expandable to 4,000 t/yr, though EF's own release gives no figure).
# Arnold Magnetic Technologies
Rochester NY. SmCo (RECOMA) = integrated alloy→assembly. NdFeB = finishing/assembly. Machines & assembles BOTH chemistries. DFARS/ITAR.
# EEC / Permag
SmCo integrated + NdFeB finishing
Electron Energy Corp (Lancaster/Landisville PA); only US SmCo producer. In-house SmCo alloy→press→sinter→machine→assemble; ALSO machines/assembles NdFeB. Sep 2026: 'more than doubles' SmCo capacity announced Aug/Sep 2025 (new alloying/pressing/fab equipment) — completion NOT confirmed by any 2026 source; mid-2026 DFARS-2027 compliance target announced. Weakest-sourced item — worth a direct ask.
# Dexter & MCE / Permag
buy blocks
Permag sisters; buy NdFeB & SmCo blocks → machine/magnetize/assemble.
Source: Permag
# Integrated Magnetics
buys blocks
Buys NdFeB & SmCo blocks → machining/assembly.
# Adams Magnetic Products
buys blocks
Buys NdFeB & SmCo blocks → fabrication/assembly.
# TDA Magnetics
Rancho Dominguez CA · AS9100D · magnet origin unstated
TDA Magnetics LLC, 1175 W Victoria St, Rancho Dominguez, CA: privately held maker of magnet components and assemblies in sintered SmCo, sintered and bonded NdFeB, ferrite and alnico, with nickel, IVD, parylene, epoxy and phenolic coatings (company website, Oct 2026). States AS9100D and ISO 9001:2015 certification. In SAM since March 2015 (CAGE 7CMX4), self-certified small, minority-owned business. One federal contract on record: NASA purchase order, $13,964, fiscal 2026, for TDA magnets for future builds. The company does not state where its sintered NdFeB or SmCo material is made and publishes no ITAR, DFARS, capacity or revenue figures.
# Ucore — Louisiana SMC
Alexandria LA · RapidSX
Strategic Metals Complex: RapidSX separation of MREC/MREO → NdPr, Dy, Tb AND Sm (SmCo-relevant). ~9,600 t/yr TREO at full build (Machine A ~600 t + 3×3,000 t lines). DoD-funded, DPAS-rated; engineering complete May 2026; Machine A H1-2027. Feedstock deal with Vulcan (Mar 2026). Kingston ON demo plant.
# Star Group + POSCO (US) Korean parent
KR-backed · STALLED
Sep 2026: STALLED / likely superseded. Announced Oct 2023 ($220M; TX/TN/AZ; 3,000→5,000 t/yr). As of Jul 2026 Star Group is still only 'in discussion with POSCO International' about a US plant of up to 3,000 t/yr — no site, no groundbreaking, no funding update in ~3 years. POSCO Int'l's actual US vehicle is now the ReElement JV (below), which covers magnet making, with no public tie to Star Group. The 'deliveries to a N.American automaker from 2026' claim has no 2026 support. Target year removed (▸TBD).
Source: G-Enews Jul-26 KED 2023 POSCO
# LS Cable & System (US, Virginia) Korean parent
KR-backed · slipped to 2029
Korea-backed. $689M advanced-materials complex, Greenbrier/Chesapeake VA (announced Dec 2025; 86 acres; ~430 jobs) — copper rod, magnet wire, AND rare-earth magnets; feed from Australia (Lynas via LS Eco Energy). Sep 2026: timeline has moved twice — Mar 2026: permits end-Q3-2027, ops late 2027; Sep 2026 (WTKR): property closing end-2026, construction after, FULLY OPERATIONAL 2029-30. Moved ▸2027→▸2029.
Source: WTKR Sep-26 VA Business VEDP
# POSCO + ReElement (US separation JV) Korean parent
KR-backed · 'Boiler Maker' · site TBD
POSCO International (KRX: 047050) and ReElement Technologies (Indiana) signed a joint venture agreement on May 21, 2026, in Washington, D.C. (POSCO's 'Boiler Maker Project'), following a Sept 2025 MOU. POSCO International is to be majority shareholder; ReElement supplies chromatographic separation technology; split not disclosed. Planned investment US$200M: US$100M for plant, equipment and initial operations, US$100M reserved for expansion on market demand. Targets: about 3,000 t/yr separated oxides (NdPr, Dy, Tb) in 2028 (Phase 1), about 6,000 t/yr by 2030 (Phase 2); pilot output Q4 2027. Metallization, magnet making and recycling are named in scope with no capacities or dates. Feed: mined and recycled material, domestic and international. As of Sept 2026, no site had been announced and no construction reported.
Source: POSCO May-26 Metalnomist Sep-26
# Aclara — Project Dynamo (Vinton, LA) + CAP metals JV
Louisiana · SPREC to NdPr/Dy/Tb/Sm oxide + NdFeB alloy · Class 3 capex US$494M · pre-FID · 2028 target
Planned separation plant on ~82 acres at the Port of Vinton, Louisiana, to process Aclara's own super-pure rare earth carbonate (SPREC) from Carina (Brazil) and Penco (Chile). Basic engineering (AACE Class 3, Hatch), 13 Oct 2026: capex ~US$494M (US$277M in Apr 2026 scoping). Study output, life-of-project average: ~1,300 t/yr NdPr, ~200 t Dy, ~35 t Tb, ~190 t Sm, ~180 t Gd, ~1,350 t Y oxide. Air permit issued Sep 2026. Site prep targeted Q1 2027 and completion H2 2028, subject to financing. EXIM non-binding letter of interest for up to US$750M (4 Sep 2026). No FID; no signed offtake. Metals: 50/50 Aclara Metals JV with CAP; FEL 2 study (Apr 2026) 811 t/yr NdPr metal and 2,681 t/yr NdFeB alloy, capex US$149M. A 175 kg/day electrolysis demo opened in Talcahuano, Chile, Oct 2026.
# REalloys — Euclid, OH
pilot-scale RE metals and alloys · DLA, Ames Lab · all current revenue
Nasdaq: ALOY. PMT Critical Metals, acquired Mar 2025 with the rare earth assets of Powdermet and Terves: R&D and pilot-scale production of rare earth metals and alloys (Dy, Nd, Nd-Dy, Mg-Nd, Gd, Sm, Sc listed) in a leased 54,000 sq ft facility, sold to the Defense Logistics Agency, Ames National Laboratory and others. US$1.51M revenue in H1 2026; no capacity or tonnage disclosed. DLA contract (Mar 2026; up to US$1.7M over 24 months per Reuters) to design a Sm/Gd metal plant of up to 300 t/yr, design only. The company's larger plans (SRC offtake, heavy rare earth metallization, Tooele, magnets) are on its profile.
Source: 10-Q, Q2 2026 REalloys facilities PMT Critical Metals acquisition
# REalloys — Tooele Army Depot, UT
Army conditional lease selection · Dy/Tb · lease not signed
Nasdaq: ALOY. US Army conditional selection (25 Jun 2026) to negotiate an Enhanced Use Lease at Tooele Army Depot for dysprosium and terbium processing, Initial Operating Capability 2028 or sooner; REalloys says the selection does not guarantee a lease. Negotiations were scheduled to finish by mid-Sep 2026; no signed lease announced as of 14 Oct 2026. No capacity or cost published for Tooele, and construction needs NEPA review and permits. REalloys' heavy rare earth metallization facility (about 50 t/yr Dy+Tb oxide feed, commissioning Q1 2028) has no site in its Q2 2026 release; Mar-May 2026 releases had the equipment built at SRC in Saskatoon and moved to Euclid, OH.
Source: US Army, 25 Jun 2026 REalloys Tooele release Q2 2026 results
# Ramaco — Brook Mine (Sheridan, WY)
critical minerals from coal · rare earths as MREC for third-party refiners · pilot 2027
Nasdaq: METC, METCB. A metallurgical coal company recovering critical minerals from coal and the clays and shales around it at its Brook Mine; rare earths are a by-product. Since a February 2026 flowsheet change (carbochlorination), the plan is to mine and make a mixed rare earth carbonate on site for third-party refiners, with no rare earth separation by Ramaco; gallium, germanium, high-purity alumina and silica are about 75% of anticipated Brook Mine revenue and scandium 18% (July 2026 letter). Pilot plant under construction: building due Oct 2026, equipment H1 2027, full pilot operation later in 2027. Scoping-level Hatch study published Jul 2026; pre-feasibility study expected in 2027; the Hatch report estimates initial production in 2031. Resource 1,392 thousand short tons of critical mineral oxides at 498 ppm on an ash basis (Inferred, effective 30 Jun 2025). Non-binding memoranda with REalloys (carbonate), Mulberry and Indium; $200M equity raised Aug 2025; $6.1M Wyoming Energy Authority grant.
Source: Shareholder letter, Jul 2026 Brook Mine FAQs Hatch Initial Assessment Q2 2026 results Weir S-K 1300 resource $200M offering closes WEA $6.1M grant REalloys MOU Indium MOU Mulberry MOU
# Peabody — Rawhide rare earth pilot (Gillette, WY)
St. Louis MO · Rawhide pilot planned · DOE selection Jul 2026
Peabody Energy (NYSE: BTU) Wyoming Rare Earths project: planned pilot plant at the Rawhide coal mine north of Gillette, WY, to make a mixed rare earth concentrate from Powder River Basin coal. Wyoming Energy Authority recommended $6.25M (Feb 2026). DOE named Peabody one of five selectees for negotiation of a $75M coal-based pilot topic (Jul 1, 2026; individual amounts not published; other selectees CONSOL Innovations, Univ. of North Dakota, Valor Metals, American Resources). Company sampling: critical mineral oxide concentrations 428-1,669 ppm, dry-ash basis, heavy rare earths 21-28% of that (Feb 2026). No resource statement, throughput or start date published. Sep 2026: the partner deal Peabody says it is nearing is for germanium, not rare earths.
Source: DOE Jul-26 WyoFile Peabody FY25
# Cyclic Materials — Mesa, AZ (Spoke) Canadian parent
Mesa AZ · magnet Spoke · up to 25,000 t/yr input · opened Sep 2026
Cyclic Materials' first commercial-scale Spoke, opened 9 Sep 2026 in Mesa, Arizona (144,000 sq ft). It runs the company's MagCycle process, which mechanically separates rare earth magnets from whole end-of-life products; outputs are Mag-Xtract magnet concentrate plus copper, aluminum and steel. No oxide is made here: concentrate goes to the company's Hub (hydrometallurgy) sites. Nameplate input: up to 25,000 metric tons/yr of magnet-bearing end-of-life products. More than 7,000 t of feedstock had been delivered by opening; first commercial shipments to US customers were expected later in Sep 2026. Announced 23 Apr 2025 with commercial operation targeted for early 2026. Investment reported as >US$20M (company, Apr 2025) and >US$30M (GPEC, Sep 2026). Feedstock partners include ERI (Jul 2026; eight US e-waste plants).
Source: Opening 9-Sep-26 GPEC
# Cyclic Materials — McBee, SC (Spoke + Hub) Canadian parent
McBee SC · Spoke + Hub · 600→1,800 t/yr MREO · ops 2028
Planned integrated campus at Alligator Industrial Park, McBee (Chesterfield County), South Carolina, announced 29 Jan 2026: a second US Spoke (MagCycle) and the company's largest Hub (REEPure hydrometallurgy) on one site. Hub input 2,000 metric tons/yr of magnet material initially, expandable to 6,000 t/yr; output 600 t/yr mixed rare earth oxide (MREO), expandable to 1,800 t/yr (company design figures). Investment >US$82M; about 90 jobs. Operations expected 2028. Groundbreaking was moved up from Q1 2027 to Q4 2026 (27 Aug 2026); no groundbreaking had been announced as of 9 Sep 2026. State support: job development credits and a US$500,000 Set-Aside grant to Chesterfield County. Intended feed includes 100% of VAC's Sumter, SC magnet swarf under a 10-year exclusive agreement (Oct 2025).
Source: SC campus Jan-26 Groundbreaking Q4
# US Strategic Metals + Ionic RE (JV)
Fredericktown MO · NdFeB + SmCo recycling · non-binding term sheet
Proposed 50-50 joint venture between Ionic Rare Earths USA (wholly owned by ASX-listed Ionic Rare Earths) and private US Strategic Metals (USSM) to recycle NdFeB and SmCo magnets and magnet scrap into separated rare earth oxides at USSM's permitted site near Fredericktown, Missouri. Non-binding term sheet announced 3 Sep 2026, superseding a Nov 2025 MOU. Intended initial funding US$100M: US$95M from USSM plus US$5M equity split equally; the first US$5M funds FEED and demonstration-scale recycling. Ionic Technologies (Belfast) would grant a non-exclusive technology license; Ionic USA would lead technical and commercial matters, including feedstock. The JV also plans to evaluate Eu, Gd, Ho, Er, Tm, Yb, Lu and Y recovery. No capacity, FID or start date published; definitive agreements targeted by end-2026.
# Phoenix Tailings
Burlington, MA + Exeter, NH · oxide to NdPr metal and DyFe alloy
Metallizes purchased rare earth oxide concentrate (supplier not disclosed) to NdPr metal and DyFe alloy by mixed-halide molten-salt electrolysis at Exeter, NH (opened Oct 2025; nameplate 200 t/yr, actual output never published), plus a 40 t/yr pilot in Massachusetts. Separation by 'selective halogenation' is at pilot scale. US$500M Office of Strategic Capital loan is a conditional commitment (Jun 2026) for a 'Freedom Facility' at an unselected site, initial operations 2028; a US$66M DOE award is selected for negotiation. Series B totals about US$116M. The '200 kg' figure in August 2026 press coverage appears in a samarium context and is not total capacity.
Source: Exeter opening OSC loan DOE selection
# HyProMag USA
Dallas–Fort Worth, TX · HPMS recycling and magnet finishing · planned
50:50 joint venture of CoTec Holdings (TSXV: CTH) and the HyProMag group under Maginito (79.4% Mkango, 20.6% CoTec). Leases half of Ironhead Commerce Center Building 1 in Northlake, Dallas–Fort Worth (about 125,000 sq ft), occupied since 3 Jun 2026. Plan as of 13 Oct 2026: cutting and finishing of magnet blocks shipped from the UK and German plants, commissioning targeted H1 2027; Phase 1 HPMS recycling and magnet making, completion targeted mid-2028, about 550 t/yr sintered NdFeB plus 484 t/yr co-products; Phase 2 about 1,048 t/yr plus 478 t/yr (study figures, not output). Jun 2026 capex estimate about US$152M. The Nov 2024 study had targeted first revenue in Q1 2027. Financing: an EXIM letter of interest for up to US$92M, not a commitment. No offtake announced.
Source: CoTec, 13 Oct 2026 Mkango H1 2026
# Momentum Technologies
Carrollton, TX · MSX separation · commercial line ramping
Private Carrollton, Texas refiner using membrane solvent extraction (MSX), with four patents licensed from Oak Ridge National Laboratory. The demonstration plant opened Oct 2024 on battery black mass and began processing rare earth magnet swarf and mining tailings in Oct 2025. On Oct 1, 2026 the company said its first commercial rare earth line was operating under an executed tolling agreement with an unnamed hyperscale technology customer, processing recycled hard-disk-drive material and ramping toward 300 t/yr (feed or product basis not stated). A second Carrollton plant of up to 1,000 t/yr in two phases is nearing FID. Company-reported purities, Jun-Oct 2026: NdPr oxide 99.5-99.9%, Dy 99.5%+, Y 99.5-99.9%. A ~2,000 t/yr Ohio plant is for battery materials, not rare earths. Investors include Freestone HoldCo (Tailwater Capital) and TechMet.
# Critical Materials Recycling — Boone, IA
Boone, IA · acid-free dissolution · HDD magnet recycling
Private Boone, Iowa recycler using acid-free dissolution recycling (ADR), invented at Ames National Laboratory's Critical Materials Innovation Hub and exclusively licensed from the Iowa State University Research Foundation. Feed: shredded hard disk drives, magnet swarf, wind turbine generators and other end-of-life magnets; product is rare earth oxide. Apr 2025: pilot with Western Digital, Microsoft and PedalPoint processed about 50,000 lb of shredded data-center HDDs and related material, with about 90% recovery of rare earth and elemental materials. A 2025 abstract described a pilot plant in commercial production (over 20 t rare earth oxide) and a demonstration plant under construction (over 100 t), time basis not stated. Aug 2025: CEO said rare earths recovered from more than 1 million drives and Boone expansion imminent. DOE Wind Turbine Materials Recycling Prize winner, Sep 2024.
Source: Western Digital release, 18 Apr 2025 TechConnect World 2025 abstract
# REEcycle (Rare Resource Recycling) unverified
Duncan, OK · magnet-to-oxide recycling · SPAC merger pending
Recovers Nd, Pr, Dy and Tb oxides from end-of-life NdFeB magnets, including disassembled hard disk drives, by hydrometallurgy based on University of Houston research. A $5.1M Defense Production Act Title III award (Jan 2025) funds restarting a demonstration plant and commissioning a commercial plant; $4.3M remained in Jun 2026. Jun 2026: a demonstration plant designed for 6-8 t/yr rare earth oxide was being commissioned at a co-location site in Oklahoma; first commercial plant targeted at 100 t/yr by 2027 (DoD gave ~50 t/yr in Jan 2025), DRA Global engineering study due Q2 2026. Principal office now Duncan, Oklahoma. Definitive agreement (May 31, 2026) to merge with Nasdaq SPAC Hall Chadwick Acquisition Corp at $400M equity value; S-4 filed Oct 1, 2026; outside date Dec 31, 2026. S-4: no material revenue to date.
Source: HCAC Form S-4, 1 Oct 2026 HCAC/REEcycle release, 1 Jun 2026
# Korea Zinc + Alta Resource — PedalPoint Korean parent
US · magnet-to-oxide recycling JV · targeted 2027
Planned U.S. joint venture of Korea Zinc (KRX: 010130) and Colorado-based Alta Resource Technologies, announced Jan 13, 2026, to recover Nd, Pr, Dy and Tb oxides from end-of-life permanent magnets at an existing U.S. site run by PedalPoint, Korea Zinc's wholly owned U.S. recycling subsidiary (site state not named). Initial phase about 100 t/yr; Korea Herald describes this as waste magnet input, Charged EVs as oxide output. Commercial operation targeted 2027, expanding in stages. Alta's process uses engineered proteins to separate rare earths. Ownership split, investment and definitive JV terms not disclosed; in Aug 2026 Alta described the JV as one the partners aim to set up. Separate from Korea Zinc's planned ~$7.4B Clarksville, Tennessee smelter (13 materials, first metal targeted Q1 2029).
Source: Korea Herald, 14 Jan 2026 Mining Weekly, 31 Aug 2026
# Bunting / Magnet Applications — DuBois, PA
DuBois PA · bonded only
Bonded NdFeB only — compression bonded and injection molded at DuBois, PA. Named by Commerce/BIS in the Section 232 NdFeB report as one of three US bonded-NdFeB producers, with Electrodyne and Tengam. Does NOT sinter: the company's own sintered-NdFeB page describes sourcing from China and Japan, and it scopes its ex-China supply claim to the bonded products only. Injection molds NdFeB, ferrite and hybrid in nylon 6/12 and PPS; tooling, pressing, coating and magnetization in-house. No capacity published. Family-owned since 1959; magnet arm acquired 2008 with Magnet Applications Group Ltd (Berkhamsted). Newton KS and Redditch UK are the separation/metal-detection equipment business, not magnet making. The atlas covers sintered magnets only, so the row is drawn for its assemblies, which use sintered NdFeB bought in, not for its bonded magnet production.
Source: BIS 232 Capabilities
# American Rare Earths — Halleck Creek, WY
Wyoming · pre-feasibility · allanite granite
ASX: ARR; OTCQX: ARRNF; ADR: AMRRY. Allanite-bearing granite at Halleck Creek on Wyoming state land. JORC resource 2,626.5 Mt at 3,292 ppm TREO at a 1,000 ppm cut-off (Jan 2025); no ore reserve. Scoping study (2024, updated Feb 2025): 3 Mt/yr, 1,833 t/yr NdPr oxide, capex US$456M, NdPr priced at US$91/kg. Pre-feasibility study targeted end-2025, then late Q3 2026, now in 'review and optimization' with no date; first production moved from 2029 to 2031. EXIM letter of interest (US$456M, Sep 2024) is non-binding. Pilot plant targets separated oxide in late Q2 2027. No binding offtake.
# Solcoa — Alameda, CA
NdPr + Sm metal · pilot claimed 10 t/yr · 500 t/yr planned
Venture-backed metal maker: NdPr and samarium metal from purchased oxide in modular reactors it builds itself, using a process it calls halide-free (no process details, purity or oxygen specifications published). Claims more than 10 t/yr from its Alameda pilot. $75M financing on 24 Sep 2026 ($45M equity led by Bain Capital Ventures; $30M debt and equipment financing anchored by J.P. Morgan) for 'Solcoa One' in Nevada, 500 t/yr, targeted July 2027. Feed reported from France, the US and Malaysia. No named customers.
Source: Sep 2026 release Axios
# Gadolyn — Kingston, TN unverified
oxide-to-alloy claim · demonstration stage
Claims a plasma-assisted, one-step process ('DIRECT') that reduces mixed rare earth and transition-metal oxides straight to NdFeB, SmCo and FeDy master alloys, with a US demonstration facility (address not published). 2025 R&D 100 award. No capacity, funding, customers or patents in the company's name found. Team from DOE's Critical Materials Institute, ORNL and Eck Industries. The one outside endorsement comes from the chair of the company's own advisory board.
Source: Advisory board release R&D 100
# Radify Metals — Campbell, CA
H-plasma oxide-to-metal · lab stage
Legal name Rareform Metals Inc. (registered in California Nov 2025). Lab-stage attempt to reduce separated rare earth oxides (dysprosium and samarium first) straight to metal with cold hydrogen plasma, water as the by-product. About $3M seed (Overture, Founders Inc., Mana Ventures, Acequia Capital). Targets several kg of metal a day by end-2026 and a pilot of up to 100 kg/day, undated. No output, purity or customers published.
Source: TechCrunch Cyclotron Road
# AML (Advanced Magnet Lab) — Melbourne, FL
PM-Wire sintered NdFeB · pilot line · DLA qualification
Private. Makes magnets by its own continuous powder-in-tube process (PM-Wire): alloy powder is filled into a non-magnetic tube, compacted, shaped and, for sintered grades, sintered. Pilot line finished in 2022 with Defense Innovation Unit funding; the company describes itself as at pilot scale (Feb 2026). $2M, two-year Defense Logistics Agency contract (May 2026) to qualify US-made sintered NdFeB, N48SH and N35EH. Its April 2026 one-pager lists 20 t/yr sintered plus 200 t/yr non-sintered now, and targets up to 5,000 t/yr sintered after 2028; no expansion site announced. Ionic Rare Earths has confirmed binding sales of NdPr and Dy oxide to AML (kilogram-scale batches so far); Phoenix Tailings and Momentum are named only by AML. Also works on SmFeN, MnBi and bonded magnets. No strip casting disclosed.
Source: Advanced Magnet Lab (AML) Awarded $2 Mil AML One-Pager: Domestic Magnet Manufactu
# NioCorp — Elk Creek, NE
Nb-Sc-Ti mine + separated NdPr, Dy, Tb oxides planned
Nasdaq: NB. Planned underground niobium, scandium and titanium mine and plant; the August 2026 feasibility study adds rare earths for the first time: separated NdPr oxide 672 t/yr, Dy oxide 67 t/yr and Tb oxide 17 t/yr on average, plus two mixed carbonates, about 25% of life-of-mine revenue. Capex $1,849M; 35 months of development from financing; no first-production year stated. Reserve 45.9M tons at 3,446 ppm TREO. EXIM support is non-binding (letter of interest, preliminary project letter, technical review under way). Rare earth offtakes with Stellantis and Traxys are non-binding term sheets. Only a mine portal is under construction; no investment decision. Where separation happens is not described in the sources read.
Source: NioCorp Project to Expand Production to 2026 Feasibility Study Results (webcast
# Chemours — Offerman, GA (monazite co-product)
mineral sands · only US monazite producer (10-K)
NYSE: CC. Mines titanium mineral sands in Florida (Starke, with mineral separation) and Georgia (Jesup and Nahunta), with mineral separation at Offerman, Georgia, the plant named in the Energy Fuels supply agreement; its FY2025 10-K calls it the only domestic producer of monazite. Tonnage, grade and buyers not disclosed. Energy Fuels has bought Chemours monazite since 2021, first under a December 2020 three-year agreement for at least 2,500 tons/yr; no renewal announced, and a March 2025 'strategic alliance' carried no terms. No rare earth cracking or separation plan found. Restructured mining in January 2026 (a mine idled, earth-moving moved to contractors).
Source: Chemours Form 10-K for FY2025 (investor- Energy Fuels Inc. Form 8-K (Chemours mon
# Rare Earth Salts — Beatrice, NE
electrochemical separation · Tb from lamp phosphor · demo plant
Private Nebraska LLC (founded 2012). Separates rare earths by a patented electrochemical process without solvent extraction at a plant it calls a commercial demonstration plant: terbium, yttrium and europium oxides from recycled fluorescent-lamp phosphor, and La, Ce and NdPr from a light rare earth concentrate. Output stated only as hundreds of kilograms of NdPr (2023), 'many metric tons' of lanthanum and one 17.5 t lanthanum sale; no capacity published. $4.22M DoD award (Sep 2024) to expand terbium oxide. A University of North Dakota-led project would add refining and metal making at Beatrice (about 21 t/yr of Nd, Pr, Tb and Dy oxides fed to metallization, about 20 t/yr of metals and alloys, from North Dakota lignite-derived carbonate); DOE's 1 Jul 2026 selection of a UND demonstration project for award negotiations does not name RES, is not a funding commitment, and the project's Feb 2026 environmental finding includes refining at Beatrice.
Source: Rare Earth Salts Announces Commercial Sc Department of Defense Awards $4.22 Milli NETL draft environmental assessment NETL FONSI, Feb 2026 DOE selections, 1 Jul 2026
# Ionic Mineral Technologies — Silicon Ridge, UT
halloysite clay · rare earths a PEA co-product · no grade published
Private, Provo, Utah. Makes nano-silicon battery anode material (Ionisil) and boehmite alumina (IonAl) from halloysite clay; neither is a magnet material. Since Dec 2025 it has presented its Silicon Ridge lease (Lake Mountains, Utah County; about 13,000 acres of state trust land) as a rare earth and critical mineral deposit. PEA (Jul 2026, S-K 1300, SGS): about 330 Mt resource, three-quarters Inferred; 2.0 Mt/yr of halloysite concentrate; 13 rare earth oxides valued as co-products, alumina about 51% of value. No rare earth grade, annual output, product form or separation step published; full report only under NDA. No offtake, government award or investment decision. Not related to Ionic Rare Earths (ASX: IXR) or Ionic Technologies (Belfast).
Source: Silicon Ridge PEA release (Jul 2026) Rare earth discovery release (Dec 2025)
Canada 9 organizations
# SRC Rare Earth Processing Facility
Saskatoon · Crown corp · REalloys offtake
Saskatchewan Research Council, a provincial Crown corporation, is building an integrated monazite processing, solvent extraction (400+ SX cells) and metal smelting plant in Saskatoon, designed for 3,000 t/yr of monazite concentrate and about 400 t/yr of NdPr metal. Smelting has run since 2024 on client oxides: about 20 t toll-produced by Dec 2025, on 40 t/yr of furnace capacity. Capital budget C$216.5M (province C$187M, SRC C$16M, NRCan C$13.5M; Apr 2026). SRC targets substantial completion Sep 2026, commissioning by Dec 2026 and ramp-up in 2027; operation was first targeted for end-2022. A five-year offtake (Dec 2025) gives REalloys most output: SRC's CEO says 70%, rising to 80% after a REalloys-funded expansion to about 600 t/yr; REalloys says about 80% of a targeted 525 t/yr NdPr metal plus Dy and Tb oxides, with intake from Q3 2027.
# Torngat Metals — Strange Lake + Sept-Îles
QC/NL · heavy-RE oxides · pre-FID · Cerberus-backed
Strange Lake, Nunavik (Quebec): planned open-pit mine and concentrator, seasonal road of about 170–180 km to a port at Voisey's Bay, Labrador, and a separation plant at Pointe-Noire, Sept-Îles. Company target about 15,000 t/yr rare earth oxides; the Quebec project notice (Aug 2025) gives 2,800–3,200 t/yr separated oxides plus 12,000–17,000 t/yr mixed carbonates. CEO-stated output (Feb 2026): 2,400 t NdPr, 540 t Dy, 80 t Tb per year. Capex about C$2B (separation plant about C$800M). Cerberus invested US$50M (2022); EDC C$110M bridge loan and CIB C$55M loan (June 2025) fund pre-FID work. Feasibility study completed March 2026, figures unpublished. Carester 10-year offtake (Oct 2025); VAC and Schneider MOUs non-binding. No FID; first production guided 2029–30. CEO Michael McCann from Oct 13, 2026.
Source: Torngat Oct-26 Mining.com Feb-26
# REalloys — Hoidas Lake
SK · S-K 1300 resource (2024) · no study, no 2026 work
Nasdaq: ALOY (REalloys). Held through Strategic Metals Development (bought May 2024, carried at US$50.5M): 14 claims, 12,522 ha, about 55 km northeast of Uranium City, Saskatchewan. S-K 1300 technical report summary (Micon, Dec 2024) includes a mineral resource estimate; no reserve or economic study. Great Western Minerals Group drilled 2001-2008; no field work from 2012 until a Sep 2024 airborne survey, and none announced in 2026. REalloys releases cite 2,153,000 tons of TREO Measured and Indicated; the Saskatchewan Mineral Deposit Index lists historical estimates of about 2.6-2.8 Mt grading about 2% rare earths. Recycling: separation trials on recycled mixed oxide at SRC (H2 2026) and a non-binding acid-mine-drainage MOU; no REalloys recycling plant.
Source: S-K 1300 report (Micon) Sask. Mineral Deposit Index 10-Q, Q2 2026
# Ucore — Kingston ON (demo + planned Sm/Gd plant)
RapidSX demo since Dec 2023
Under consideration for commercial scale. 52-stage RapidSX demonstration plant (~5,700 hrs, 99.5% NdPr shown; heavy MREO processed). Government of Canada conditional approval up to C$36.3M (NRCan C$26.3M + FedDev C$10M, 31 Oct 2025) for a commercial samarium/gadolinium refining plant in Kingston — no capacity/date. Ucore's commercial build is Louisiana (US section, ▸2027, DoW OTA US$22.4M).
Source: Ucore Oct-25 Ucore LA
# Cyclic Materials
Kingston ON · Hub100 mixed oxide · Solvay supply · 500 t/yr hub unconfirmed
Founded 2021 in Kingston, Ontario (HQ now Toronto). Kingston hosts the company's REEPure hydrometallurgy Hub: Hub100 opened Jun 2024 with design capacity to recycle 100 metric tons/yr of magnet material into recycled mixed rare earth oxide (rMREO) plus nickel-cobalt hydroxide; a mechanical pilot line (Dec 2023) has 8,000 t/yr design input. A US$25M Center of Excellence (140,000 sq ft; 500 t/yr magnet-rich feedstock to rMREO; R&D labs) was announced 11 Jun 2025 for a Q1 2026 start; no opening announcement found as of Oct 2026. Individual oxides are not separated on site: a Jun 2024 supply agreement sends rMREO to Solvay, La Rochelle, France, and a Mar 2026 non-binding MOU covers supply to Neo. Grants: NRCan CAD 4.9M (Sep 2024). Total equity raised US$237M (Aug 2026).
Source: Hub100 Jun-24 Solvay supply
# Geomega Resources (Innord)
Saint-Hubert, QC · TSXV: GMA · magnet recycling demo
Geomega's subsidiary Innord is building a demonstration plant in Saint-Hubert, Quebec, to recover the main rare earth oxides found in magnets from end-of-life and scrap NdFeB magnets. Design feed is 1.5 t/day of magnet material (company presentation, June 2026: about 500 t/yr feed, 150 t/yr REO). Quebec's environment ministry authorized operation and data collection through Dec 31, 2028 (June 2026); commercial operation would need a separate permit. As of Sep 2, 2026, process piping was more than 75% complete and commissioning of process equipment was expected to begin in November 2026; production was first targeted for 2020. No oxide offtake or scrap supply contract has been announced. Funding includes Quebec and federal grants and a C$1.35M secured loan (Sep 2026).
# Mont Royal — Ashram, QC Australian parent
Nunavik carbonatite · concentrate + Saguenay MREC plant · PEA
ASX: MRZ; TSX-V: MRZL. Acquired Ashram with Commerce Resources (closed Oct 2025). Monazite-dominant carbonatite in Nunavik, about 130 km south of Kuujjuaq, with no road access. Resource 73.2 Mt Indicated at 1.89% TREO plus 131.1 Mt Inferred at 1.91%; NdPr about 21% of TREO. Updated PEA (June 2026): 1.8 Mt/yr ore to a flotation concentrate of about 30% REO on site, then a proposed hydrometallurgical plant at Saguenay producing mixed carbonate with about 4,035 t/yr NdPr contained; capex C$1.231B excluding a road of about C$684M; no production year. A PFS promised for 2013-14 was never published; a new PFS is targeted to start H2 2026. Federal road-study support (up to C$2.6M) is conditional.
Source: Updated Preliminary Economic Analysis Co Mont Royal Files PEA Technical Report fo
# Defense Metals — Wicheeda, BC
carbonatite · concentrate + MREC plant · PFS
TSX-V: DEFN. Wicheeda carbonatite about 80 km northeast of Prince George, 100% owned since 2022. PFS (Feb 2025, US$): 5,000 t/d concentrator to a 50% TREO concentrate, and an off-site hydrometallurgical plant at Bear Lake making mixed rare earth carbonate containing 4.4 kt/yr NdPr oxide equivalent; capex $1.44B; NdPr priced at about $133/kg; first reserve about 26 Mt at 2.4% TREO. Feasibility study guided to start in early 2026, not yet started as of September 2026. Support all non-binding or conditional: EDC letter of interest up to US$250M, conditional NRCan grants, non-binding MoUs including Hanwha. Equity partnership with the McLeod Lake Indian Band.
Source: Defense Metals Announces Positive Prelim Defense Metals Corp. Acquires 100% of th
# Search Minerals — Deep Fox and Foxtrot, Labrador
Labrador deposits · MREC planned · 2022 PEA
TSX-V: SMY. Deep Fox and Foxtrot, Port Hope Simpson–St. Lewis district, southeast Labrador. NI 43-101 resources (Dec 2021): Deep Fox 5.05 Mt Indicated plus 3.30 Mt Inferred; Foxtrot 10.04 Mt Indicated plus 3.01 Mt Inferred, grades reported by element only. 2022 PEA: 2,000 t/day ore to a mixed carbonate by the company's patented direct extraction process, separation tolled; processing plant sited on the Island of Newfoundland; about 1,437 t/yr magnet rare earth oxides; initial capex C$422M; production goal of 2025 not met. Environmental assessment registration expected H1 2027. A 2021 non-binding MoU with USA Rare Earth (500 t/yr Nd/Pr); no later statement says whether it still stands. Cease trade order 2024-2025; market cap about C$12M (Sep 2026).
Source: Search Minerals Announces Positive Preli Search Minerals Launches 2026 Environmen
United Kingdom 4 organizations
# Less Common Metals (LCM) US-owned
Ellesmere Port · USARE subsidiary
RE metals + strip-cast NdFeB alloy (+ SmCo alloys); first commercial yttrium metal pour Apr 2026. Fastmarkets reports ~2,500 t/yr total alloy capacity operating (Sep 2026), across light/heavy RE metals, magnet alloys, hydrogen-storage alloys and other specialty materials; the previously stated 3,000 t/yr Q4-2026 run-rate target did not reappear in USARE's Q2 2026 outlook. Largest ex-China alloy plant in Europe; direct competitor to Santoku. Recycling-based sintered makers HyProMag (Birmingham/Pforzheim/DFW) and Ionic (Belfast, oxide only) on the bench. Recycling: in the REEmelt project LCM purifies and remelts recycled magnet alloy directly, described as the only UK company able to do that type of remelting at scale. Note the contrast with the Ionic/VAC chain, where LCM only reduces oxides that Ionic recovered. No tonnage published.
Source: REEx USARE Fastmarkets
# Ionic Technologies (Belfast)
ASX:IXR · 10 t/yr REO demo · 400 t/yr plant pre-FID
Wholly owned Belfast subsidiary of Ionic Rare Earths (ASX: IXR); Queen's University Belfast spin-out founded 2016, acquired 2022. Demonstration plant rated 10 t/yr separated REO; supplied 120 kg Nd2O3, 10 kg Dy2O3 and 8 kg Tb4O7 (100% recycled) for a Ford UK e-motor test with LCM and GKN (May 2026); binding oxide sale agreements with US-based Advanced Magnet Lab, first batch June 2026, volumes undisclosed. Planned commercial plant at Queen's Island: 1,200 t/yr magnet and swarf input, 400 t/yr separated Nd, Pr, Dy and Tb oxides (Nov 2024 feasibility study), capex GBP 85M. The GBP 12M DRIVE35 grant is an offer in principle; the rest is unfunded. FID was targeted by end-Sep 2026 and has not been announced; planning application lodged 19 Sep 2026, decision expected Q1 2027; construction targeted Q3 2027.
# HyProMag Ltd
Tyseley, Birmingham · HPMS recycled alloy and sintered blocks
Operated by HyProMag Ltd (100% Maginito: 79.4% Mkango, 20.6% CoTec) in a facility at Tyseley Energy Park built by the University of Birmingham. Uses the university's patented Hydrogen Processing of Magnet Scrap (HPMS) process, licensed exclusively to HyProMag. First recycled NdFeB alloy powder in Jul 2025; officially opened 15 Jan 2026 by UK industry minister Chris McDonald. As of 1 Sep 2026: 10.1 t of alloy powder produced in total, presses and sintering furnaces commissioned, and small-scale sintered blocks made for customer evaluation. Machined parts come from a machining partner. Stated capacity is 100 t/yr of sintered magnets on one shift and over 300 t/yr on multiple shifts; steady state targeted end-2026. A £1.12M feasibility study (60% DRIVE35 grant) covers expansion to 1,000+ t/yr. Siemens and JLR trials; no binding offtake.
Source: Mkango H1 2026 Opening, Jan 2026
# Bunting / Magnet Applications — Berkhamsted
Berkhamsted · assemblies
Bespoke magnetic assemblies and industrial magnetizing equipment at Berkhamsted, Hertfordshire, the European head office of the magnet business (Bunting Magnetics Europe Limited, renamed January 2012); machining, testing and warehousing added at Swindon with the MagDev acquisition of October 2021. No magnet forming is described at either UK site. Magnet production is bonded NdFeB at DuBois, PA - see that row and the company profile, which also records an ambiguous 2018 grain boundary diffusion claim.
Germany 4 organizations
# Vacuumschmelze (VAC)
Hanau; VACODYM (NdFeB) + VACOMAX (SmCo), 170+ alloys; melt→sinter→machine→assemble. Sites: Hanau DE, Horná Streda SK (permanent magnets), Ulvila FI (= Neorem Magnets, sintered NdFeB — own row under Finland), Sumter SC (eVAC, live Dec 2025). Sep 2026: Energy Fuels acquisition ($1.9B) agreed Jun 2026, NOT closed as of 15 Oct 2026; first amendment to the merger agreement 13 Oct 2026; Energy Fuels' statements on timing differ ('early 2027' vs within 12 months of the June signing). Energy Fuels' 8-K (13 Oct 2026) gives site capacities for the first time, as the acquirer's figures: Hanau permanent magnets 1,000 t/yr, Ulvila 500 t; VAC itself publishes none. JV w/ LS (Korea); alliance w/ Aclara (Chile/Brazil/LA) for ex-China feed; Cyclic Materials recycling deal (10-yr exclusive on Sumter NdFeB swarf, Jan 2026). Context: GKN Powder Metallurgy CANCELED its planned 4,000 t/yr European NdFeB plant (Dec 2025).
Source: VAC EF deal GKN exit Energy Fuels 8-K, 13 Oct 2026 (VAC capacities)
# Magnetfabrik Schramberg
buys alloy
Buys alloys. Bonded/pressed NdFeB + SmCo (+ferrite excluded). Starts at block making.
# Remloy — Bitterfeld
Bitterfeld-Wolfen · melt-route NdFeB alloy powder · Mkango since Aug 2026
Melt-based recycling of end-of-life NdFeB magnets and rare earth alloys into NdFeB alloy powder for bonded and hot-deformed magnet makers, at a plant in Bitterfeld-Wolfen, Germany. Heraeus announced production start in May 2024 (stated then as up to 600 t/yr, expandable to 1,200 t/yr). Mkango Resources agreed on May 20, 2026 to buy the business from Heraeus Amloy Technologies for EUR 8M cash and announced completion on Aug. 31, 2026 (EUR 5M paid; EUR 3M due Aug. 28, 2028). Mkango gives a target of at least 500 t/yr of alloy powder over the next few years; 2026 output is small-scale (samples, product optimization, initial sales), with first commercial sales targeted by end-2026. The business was non-revenue as of May 2026. About 345 t of magnet and alloy stockpile transferred; no binding offtake reported.
Source: Mkango completion Heraeus launch 2024
# HyProMag GmbH
Pforzheim · HPMS recycling, commissioning · permitted to 750 t/yr
HPMS short-loop recycling of NdFeB scrap into alloy powder and sintered magnet blocks. About 90% held by Maginito (79.4% Mkango, 20.6% CoTec), assuming conversion of a Maginito convertible loan. Officially opened 28 Apr 2026 by Germany's Federal Ministry for Economic Affairs and Energy. As of 1 Sep 2026: HPMS vessel, jet mill and transverse and axial presses commissioned and first pressed blocks made; remaining equipment due in the coming months; a 1,000 kVA transformer and 20,000-liter liquid nitrogen tank installed. Permitted for up to 750 t/yr. Stated minimum initial capacity is about 100 t/yr, about 350 t/yr on multiple shifts, with 750 t/yr under evaluation over a three-year phased ramp (permit and capacity figures, not output). Partly funded by the ERDF and Baden-Württemberg. No binding offtake announced.
Source: Mkango H1 2026 Opening, Apr 2026
Poland 1 organization
# Mkango — Pulawy separation plant, Poland Canadian parent
planned SX plant · 30,000 t/yr MREC feed
Planned solvent-extraction plant on an 8-hectare site leased from Grupa Azoty Pulawy, to separate Songwe's mixed carbonate: 30,000 t/yr of carbonate feed (13,430 t/yr REO) to about 3,037 t/yr Nd oxide and 920 t/yr Pr oxide. Pre-feasibility study (Mar 2026); commissioning targeted Oct 2028, ramp-up 2029-2030. Pre-investment decision. EU strategic project. Capex given as about US$212M in the study and US$128.2M in a June 2026 SEC-filed presentation.
Switzerland 1 organization
# Arnold Magnetic Technologies AG (Lupfig) US-owned
RECOMA SmCo · European Arnold site
Arnold's European SmCo operation (ex-Ugimag), Lupfig AG: RECOMA Sm2Co17/SmCo5 magnets — pressing, sintering, machining, finishing and assemblies; drawn Alloying→Assembly to match Rochester (in-house alloying at Lupfig to be confirmed). Arnold's Thailand plant (Chonburi, Phase II mid-2027) is being built partly to relieve Lupfig. Switzerland is NOT a DFARS qualifying country (7052-OK if feed is non-China); UK (Sheffield) is. Suggested by a reader of the map (Sep 2026).
Source: Arnold sites RECOMA
Finland 1 organization
# Neorem Magnets (VAC) German parent
Ulvila · buys alloy · large sintered NdFeB
Sintered NdFeB specialist for LARGE magnets, pole elements and assemblies (wind generators, big motors, servo/linear): hydrogen decrepitation → jet milling → die pressing in field → sintering → machining → coating → magnetizing → assembly, all in Ulvila; starts from purchased alloy (no melting/strip-casting disclosed). Ex-Outokumpu Magnets (1988; NdFeB under license since 1991; Neorem 1995); part of Vacuumschmelze since 2007 — so VAC's 'Ulvila' site IS Neorem (corrects an earlier note that it was soft-magnetic). Finland = DFARS qualifying country. Capacity: 500 t of permanent magnets per Energy Fuels' 8-K (13 Oct 2026), the acquirer's figure with no annual qualifier; Neorem publishes none. Suggested by a reader of the map (Sep 2026).
Source: Neorem Production Energy Fuels 8-K, 13 Oct 2026 (VAC capacities)
Sweden 2 organizations
# Leading Edge Materials — Norra Kärr Canadian parent
heavy-rich eudialyte deposit · mining concession granted 2026
TSX-V: LEM; Nasdaq First North. Eudialyte-hosted deposit in Jönköping County, rich in heavy rare earths. The Swedish government granted a 25-year exploitation concession on 28 June 2026 (guided for 2025); an environmental permit is still needed before operation. An earlier lease (2013) was revoked by the Supreme Administrative Court in 2016. Latest study is a 2021 PEA on an Inferred-only resource of 110 Mt at 0.5% TREO (heavy rare earths 54%): 5,341 t/yr TREO as mixed oxide including 248 t/yr Dy oxide and 36 t/yr Tb oxide; capex US$487M. A new PFS is guided for 2026. Not selected as an EU strategic project in March 2025. Working capital C$0.54M at April 2026 with a statement that funds were insufficient for 12 months, before a C$4.28M placement.
Source: Tasman announces results of Pre-Feasibil Tasman Minerals mining license revoked b
# LKAB — Luleå demonstration plant (apatite to rare earth concentrate)
state iron-ore miner · rare earths from apatite in tailings · demo plant
Swedish state-owned iron-ore miner. A demonstration plant at Svartön, Luleå (about SEK 800M, approved Oct 2024) treats apatite concentrate from Malmberget iron-ore tailings to phosphorus products, gypsum and a mixed concentrate of all 17 rare earths; LKAB said in May 2026 it would be in operation in autumn 2026, with no start announced as of late September. No capacity published. Separation into individual elements is planned with REEtec in Norway, in which LKAB invested NOK 400M in 2022, becoming the largest shareholder (34.4% at end-2025); no offtake contract published. Environmental permit for the full-scale Luleå industrial park granted 26 May 2026; no investment decision, full scale now targeted for the 2030s (2027 in 2022 plans). Per Geijer at Kiruna (Inferred only, about 2.2 Mt REO) is the long-term source. EU strategic projects (Mar 2025).
Source: LKAB constructs facility for critical mi LKAB receives the go-ahead for its criti LKAB annual report 2025
France 5 organizations
# Solvay
La Rochelle · the hinge · Sm since 2025, Dy/Tb autumn 2026
La Rochelle has separated rare earths since 1948 - the largest such plant outside China and one of very few able to process both light AND heavy REs at industrial scale (FY2025 group underlying net sales EUR 4.3bn; RE sits in Special Chem under An Nuyttens). Apr 2025: new magnet-grade NdPr line (~EUR 10M). 2025: SAMARIUM PRODUCTION STARTED (2025 Annual Integrated Report). Nov 2025: agreement with PERMAG (US) + LESS COMMON METALS (UK) - Solvay separates pure Sm oxides at La Rochelle, LCM metallizes, Permag makes magnets/assemblies, 3-5 year horizon. Jul 2026: up to EUR 20M into heavy-RE separation, industrial-scale Dy/Tb from autumn 2026 (verify Sep-2026). Target: 30% of European magnet-grade light+heavy RE demand by 2030; -40% CO2, -50% water. Feedstock network: Cyclic Materials rMREO offtake (2024), Viridis Mining & Minerals LoI (Jun 2026, Brazilian ionic clay by 2028 - carries Sm/Gd/Y as well as Nd/Pr/Dy/Tb), Carester MOU (2024). Open to a US plant but conditions it on stronger government/customer commitments - nothing greenlit. Separation only, no magnets. Recycling: a non-binding 2023 MoU with Cyclic Materials to separate recycled mixed rare earth oxide shipped from Cyclic's Ontario site to La Rochelle, plus a stated aim to play a proactive role in magnet recycling. No volumes, no timeline, and no confirmation the MoU has advanced.
Source: Solvay Permag+LCM Sm Viridis LoI EUR 20M Dy/Tb Profile
# Carester / Caremag
Lacq, France · magnet recycling · NdPr + Dy/Tb oxide separation
Lyon-based Carester (founded 2019) is building Caremag at Lacq, France: design input 2,000 t/yr end-of-life magnets plus 5,000 t/yr mining concentrates; design output at full ramp 800 t/yr NdPr oxide (650 t didymium + 150 t Nd oxide), 500 t/yr Dy oxide and 90-100 t/yr Tb oxide. FID Dec 2024; construction since Apr 2025. Financing €216M (Mar 2025): €106M French state, up to €110M from the JOGMEC/Iwatani JV. Start guidance: operations Q4 2026 (USA Rare Earth, Jul 2026); operations targeted 2027 (Lindian, Sep 2026). Binding offtake: Stellantis, 10 yr, >3,400 t oxides (Sep 2024); long-term heavy-RE supply to Japan. Binding feed offtakes: Brazilian Rare Earths (Oct 2025), Lindian (Sep 2026, conditional). USA Rare Earth and InfraVia signed definitive agreements for ~13.6% each (Jul 2026); closing not yet announced. No metal or magnet production.
Source: Carester Mar-25 USARE Jul-26
# USA Rare Earth — LCM Europe US-owned
Lacq · metal + alloy · announced Jan 2026
LCM Europe SAS: a USA Rare Earth subsidiary announced 20 Jan 2026 to build a rare-earth metal and complex-alloy plant at Lacq, targeting approximately 3,750 t/yr across light and heavy rare-earth metals and complex alloy systems. A separate legal entity and facility from Less Common Metals Ltd. (LCM UK, Ellesmere Port) despite the shared ownership, and co-located with but distinct from Carester's neighboring Caremag oxide plant (see that entry). French government support: C3IV program credits for up to 45% of eligible equipment costs, plus up to EUR 130M toward real estate; USARE/LCM's own figure for total project cost is EUR 110M, with an estimated 100-140 jobs. No construction start, groundbreaking, or commissioning date has been published for this facility specifically as of Sep 2026.
Source: USARE 20-Jan-26 Fastmarkets
# MagREEsource
Noyarey, Isere · magnet-to-magnet recycling · 50 t/yr sintered NdFeB
CNRS-Institut Neel spin-off (founded 2020) that recycles sintered NdFeB magnet scrap by hydrogen processing and makes new sintered magnets from it at Noyarey, near Grenoble. The plant was inaugurated June 27, 2024 (EUR 8.5M invested). A May 2026 company presentation gives 50 t/yr of NdFeB magnets for 2026-2028, grades up to N42SH from fully recycled feed, a SmCo pilot line and 55 staff. A melt-based 'short-loop fusion' step is being added (equipment expected 2026). Raised EUR 23M in July 2025, led by Bpifrance's SPI 2 fund (France 2030). MagFactory, an EU Critical Raw Materials Act Strategic Project (Mar. 2025), targets 1,000 t/yr of magnets; timing stated as 2028 (2025) and 2028-2031 (May 2026). No MagFactory construction financing announced as closed.
# Orano + CEA Liten — Grenoble magnet pilot line
short-loop recycled sintered NdFeB · ~4 t/yr pilot
Pilot line run by Orano and CEA Liten on CEA premises in Grenoble, inaugurated 19 Sep 2025 (about 220 m2), funded by France Relance/Bpifrance (Magnolia consortium with Valeo, Paprec and others) and Horizon Europe (MAGELLAN, coordinated by Orano, EUR 6.50M EU contribution, to Mar 2027). Short-loop recycling: end-of-life magnets remelted to alloy without separating the rare earths, strip-cast, jet-milled, pressed in field and sintered; target 25% recycled rare earth content. Capacity given as 4 t/yr of magnets (one source gives 50 kg/day, which would be 18 t over a full year). Reproducibly made grade N45UH in Sep 2026. Orano says it will transfer the process to an industrial partner rather than build a plant; no partner named. State-owned (French state 90.33%).
Source: CEA, Sep 2026 Orano, Sep 2025 Interesting Engineering
Norway 1 organization
# REEtec unverified
Herøya · NdPr oxide · start-up pending · LKAB 33.7%
Separation plant at Herøya Industrial Park, Porsgrunn, Norway, using a proprietary process in development since 2008; a demonstration plant has run on site since 2019. Plant 1 nameplate is 720 t/yr NdPr oxide, funded by a NOK 1.2B (about EUR 115M) equity round in Nov 2022. First production was targeted for H2 2024; start-up commissioning began Aug 2024, and an April 2025 Norwegian press report said the plant would be ready for production in 2026. No commercial output reported publicly as of Oct 2026. Named feed supplier Vital Metals terminated its offtake effective Dec 26, 2023 (REEtec disputes it). Schaeffler signed a five-year oxide purchase contract in April 2022. LKAB is the largest owner (33.7%, Aug 2024); a second plant using LKAB concentrate was planned for 2026–27.
Greenland 2 organizations
# Tanbreez — Greenland
Killavaat Alannguat · eudialyte
Eudialyte-hosted HREE in kakortokite, Killavaat Alannguat, Kujalleq. Exploitation license MIN 2020-54, 18 km², 30-year term. Critical Metals Corp (Nasdaq: CRML) 92.5% since Greenland approved the final 50.5% transfer 17 Apr 2026; European Lithium (ASX: EUR) 7.5%. Only classified resource is 44.87 Mt @ 0.38% TREO (S-K 1300, Ind+Inf), effective date Aug 2016; HREE ~27% of TREO. PEA only — no feasibility study, no FID. Pilot plant (300-500 kg/hr) footings under way Jun 2026, no commissioning. First ore targeted late 2028-early 2029. Dy and Tb grades are not published anywhere.
# Greenland Mines — Sarfartoq US-owned
Sarfartoq carbonatite · Initial Assessment · Neo offtake
Nasdaq: GRML (formerly Klotho Neurosciences, a gene-therapy company, renamed March 2026 after buying Greenland Mines Corp., owner of 80% of the Skaergaard palladium-gold project). Bought Sarfartoq from Neo Performance Materials and co-owners for US$35M ($20M cash, $15M in shares), closed 1 Sep 2026; Neo keeps offtake rights on up to 60% of ore or concentrate, for its Silmet plant in Estonia, which Greenland Mines describes as non-binding. Carbonatite (bastnaesite, monazite). S-K 1300 resource, ST1 zone: 6.9 Mt Indicated at 1.60% TREO plus 5.3 Mt Inferred at 0.96% (Jul 2026); NdPr about 25% of TREO. Initial Assessment: 1.4 Mt/yr for 9 years to a flotation concentrate of about 8.25% TREO at 63.6% recovery; headline economics are a high case only. Exploration license MEL 2020-32; a 262 km2 extension has been applied for. First explored by Hudson Resources (inferred-only resources, 2011-2012).
Source: Initial Assessment Neo close Resource
Estonia 1 organization
# Neo Performance Materials
Silmet/Sillamäe = separation (small HREE SX line commissioned Apr 2026) → Narva = Europe's largest sintered-NdFeB plant (opened Sep 2025). Sep 2026: Q2-2026 still shipping QUALIFICATION samples for awarded auto platforms; 2-3 programs to enter commercial production in 2026 → real volume 2027 (mild slip). Phase 1A ~2,000 t; Phase 1B equipment ordered → ~5,000 t (~2028). Metallization/Alloying gap persists — Neo's metal plants are in China/Thailand. Bidder in India's REPM scheme. Canadian parent (TSX:NEO).
Slovenia 2 organizations
# Magneti Ljubljana
buys alloy
Buys alloys. Sintered/bonded NdFeB (+ferrite excluded). Recycling: a SUSMAGPRO pilot line rated at roughly 20 t/yr of recycled NdFeB, plus EU projects (DyScovery, MaXycle, REMANENCE, DEMETER) on end-of-life NdFeB and SmCo including WEEE, and a separate national project on recycling sludge from its own Sm2Co17 production. All framed as projects rather than a commercial production process.
# Kolektor
Idrija SI · bonded NdFeB + assemblies · sintered from CN affiliates
Kolektor Mobility (Idrija, Slovenia), part of the privately held Kolektor Group: makes compression- and injection-molded bonded NdFeB, hard-ferrite and mixed magnets, rotor and sensor magnets, stators and assemblies; company states 250 million magnets/yr (count, all materials). Bonded NdFeB is made from purchased Magnequench MQP-B+ powder (May 2026 paper). Production moved from Essen, Germany (ex-Kolektor Magnet Technology) to Idrija in 2020-21. Sintered NdFeB and SmCo magnets it sells come from affiliated companies in China, not made in-house; the atlas covers sintered magnets only, so the row is drawn for that work, not for its bonded production. June 2026: Slovenian chamber gold award for polymer-bonded magnets from 100% recycled NdFeB powder (EU HARMONY partner; validation with Copreci ongoing).
Australia 5 organizations
# Lynas Rare Earths
Mt Weld (WA) + LAMP Kuantan (see Malaysia) + Kalgoorlie. NdPr AND heavy/Sm: first Dy oxide May 2025, Tb Jun 2025, SAMARIUM oxide Mar 2026 (→SmCo); Gd/Y early 2028, Lu Apr 2028 (HRE capex blew out A$180M→A$294M on non-Chinese equipment). No magnets itself ('miner + chemical processor'): JS Link JV in Malaysia + LS Eco Energy metal JV in Vietnam. Sep 2026: TEXAS (Seadrift) = 'on hold / unlikely to proceed' — never formally canceled. Malaysia operating license renewed for a 10-year term from 3 Mar 2026 (was short-cycle renewals). FY26 sales volume 12,122t REO/7,337t NdPr (NOT production - production was 13,089t REO/7,260t NdPr; sales > production this year, drawing down inventory). Realized price A$80.7/kg REO across all products, not NdPr-specific. FY26 record revenue A$978M; JARE terms disaggregated (FY26 report): firm 5,000t/yr NdPr @ US$110/kg floor, up to 7,200t/yr available to 2038; firm 50% HRE oxides, up to 75% available (widened Mar-26 to add Sm/Y/Lu/Gd). Separately, US DoW 4-yr deal also carries a US$110/kg NdPr floor - same benchmark, two contracts, not one figure double-counted. CEO Lacaze retired; Pol Le Roux interim. Mar 2026: Sojitz/JARE revised long-term contract raises Japan's committed share of Lynas medium/heavy RE output to 75% (from 65% Dy/Tb under the 2023 deal) and adds Sm/Y/Lu/Gd; no tonnage or end date in Sojitz's own release.
Source: REEx FY26 Texas Sojitz 13-Mar-26
# Arafura Rare Earths (Nolans)
NT · ore-to-oxide · 4,440 t/yr NdPr oxide nameplate · FID May 2026 · first production 2029
Nolans, ~135 km N of Alice Springs, NT: mine, beneficiation, extraction and separation plants on one site, ore to NdPr oxide. Study/nameplate: 4,440 t/yr NdPr oxide plus 470-573 t/yr SEG/HRE oxide and ~144,000 t/yr phosphoric acid; 38-yr life. FID 21 May 2026; Hatch EPCM; first sod 8 Oct 2026. Guidance: ~37-month build, first production mid-to-late 2029, nameplate ~2031-32. Funding: US$775M senior debt conditionally approved since Jul 2024, contractual close targeted Oct 2026 (not announced by 16 Oct); A$723M cash+deposits 30 Jun 2026; EFA US$100M and GRMF EUR50M binding equity subscriptions; NRFC A$200M convertible notes. Offtake: binding Hyundai/Kia (up to 1,500 t/yr), Traxys NA (850 t/yr, Oct 2026), Traxys Europe (100-300 t/yr), wind OEM (up to 500 t/yr, extended Sep 2026), all with conditions precedent; Indian SPV term sheet and EFA reserve letter of support (500 t/yr each) not final.
# Iluka Resources — Eneabba refinery
WA · up to 23,000 t/yr REO · NdPr + Dy/Tb oxide · mid-2027
Fully integrated rare earth refinery under construction at Eneabba, WA: roasting, leaching, purification, solvent extraction and finishing to separated Nd, Pr, Dy and Tb oxides plus other light and heavy oxides; no metal or magnet production. Design feed 55,000 t/yr concentrate; capacity up to 23,000 t/yr REO, 5,500 t/yr NdPr oxide and 750 t/yr Dy/Tb oxide, subject to feedstock (stockpile-only feed: about 3,500 t NdPr and 170 t DyTb). Capex estimate A$1.7-1.8B (A$1.0-1.2B at 2022 FID); A$1.1B spent to 31 Jul 2026; construction nearing 60% (Aug 2026). A$1.65B non-recourse Australian Government loan via EFA, A$897.5M drawn incl. capitalized interest at 30 Jun 2026. Commissioning scheduled mid-2027 (2025 at FID). First binding offtake (Jun 2026): 1,200 t magnet oxides to an unnamed automaker, 2028-2031, take-or-pay, US$155M minimum.
Source: Iluka H1 2026 Iluka offtake, Jun 2026
# Tronox — Western Australia (monazite + planned MREC plant) US-owned
mineral sands monazite · 10,000 t/yr MREC planned
NYSE: TROX. A titanium dioxide producer that recovers monazite as a co-product of its mineral sands mines in Australia and South Africa and sells it to rare earth processors; the CEO told Congress in 2023 that it had historically been sold to Chinese processors. Tonnage and buyers are not disclosed; it is reported inside 'other products'. Plans a cracking and leaching plant in Western Australia making 10,000 t/yr (TREO basis) of mixed rare earth carbonate, light and heavy; PFS complete, DFS due Q3 2027, start-up expected late 2029, no site, capex or FID published. EXIM and Export Finance Australia letters of up to US$600M are conditional and non-binding (Dec 2025). A US separation refinery is being scoped as a second phase (Hamilton, MS among options). Holds about 5% of Lion Rock Minerals (Cameroon). Says it is not pursuing metals or magnets.
Source: Export-Import Bank of the United States Tronox Holdings plc Form 10-K for fiscal
# Australian Rare Earths — Koppamurra, SA
ionic clay · heap leach to mixed oxide · PFS
ASX: AR3. Ionic clay deposit near Naracoorte, South Australia. Resource 243 Mt at 751 ppm TREO; maiden probable reserve 26 Mt at 920 ppm TREO (June 2026). Pre-feasibility study (June 2026): 3 Mt/yr of ore by heap leach to a mixed rare earth oxide, 1,860 t/yr TREO containing 435 t/yr NdPr and 57 t/yr Dy+Tb; capex A$178M; first production targeted by 2029 (earlier 2025/26). About 30 t pilot at ANSTO since August 2026. Mining lease application targeted for end-2026. Only offtake found: a non-binding 2022 MoU with Neo for half of first-module output.
Source: Koppamurra Pre-Feasibility Study (ASX re Australian Rare Earths releases compelli
Malaysia — foreign-owned plants (Kuantan hub) 2 organizations
# Lynas Advanced Materials Plant (LAMP) Australian parent
Kuantan/Gebeng
Cracking/leaching + separation of Mt Weld concentrate: NdPr, La/Ce, and since 2025-26 Dy, Tb and Sm oxides — the only commercial-scale heavy-RE separation outside China. Gd/Y/Lu circuits ▸2028. Feeds JS Link (KR/MY), LS Eco Energy (VN), Noveon (US), and JARE/Sojitz (JP).
Source: Edge Q4-FY26 call
# JS Link — Kuantan, Malaysia Korean parent
Kuantan · 3,000 t/yr · planned
Planned 3,000 t/yr sintered NdFeB magnet plant in Kuantan, Pahang, Malaysia, next to the Lynas Malaysia plant, to be built and run by JS Link through a subsidiary (provisional name JS Link Magnetics Malaysia; KRW 50bn capital contribution approved 25 Mar 2026). Not a JV: Lynas holds ~4.58% of parent JS Link (A$50M, closed 29 Jul 2026) and is exclusive rare earth supplier to Jan 2038 under a 7 Jul 2026 agreement that followed a non-binding Jul 2025 MoU. Reuters (Nov 2025) put the investment at RM600M (US$142M); the RM142M figure is Lynas's equity, not plant cost. Site secured, up to 400 jobs, jet mills ordered. JS Link targets commercial operation by early 2028; no construction start announced as of Oct 2026.
Source: Lynas ASX MT/The Bell
Thailand — midstream host (no mine, no separation) 3 organizations
# Hastings + Enuo — Kabin Buri MREC plant Australian parent
Prachinburi · monazite to MREC · first output target Q1 2027
Hastings Technology Metals (ASX: HAS) signed a binding term sheet on Mar 31, 2026, to buy 49% of a permitted monazite cracking and leaching plant at Kabin Buri, Prachinburi, from Enuo Holdings (Singapore) for up to US$15M; completion was still pending in Sept 2026. Product is mixed rare earth chloride (MREC), a separation feed, not separated oxides. In Jul 2026 planned output was raised from 6,000 to about 12,000 t/yr of MREC (product tonnage) on about 10,000 t/yr of feed; a 36,000 t/yr stage is cited for Q4 2028. Feed: Enuo agreement (Apr 10, 2026) for at least 10,000 t/yr of African monazite (at least 54% TREO) delivered to Laem Chabang. Site works began Aug 2026; first MREC is targeted for Q1 2027 (was Q4 2026). No binding MREC sales agreement has been announced.
Source: ASX 31-Mar-26 Hastings 14-Jul-26
# GQD Special Material (Thailand) Chinese parent
Rayong · Neo 20% · Ganzhou Qiandong group · Nd oxide-to-metal toll
Thai joint venture registered February 14, 2013 (registered capital THB 165.79 million) with a plant in Pluak Daeng, Rayong. Neo Performance Materials owns 20% and, per its annual information forms (2020 and 2025), outsources conversion of neodymium oxide from its NPM Silmet plant in Estonia into metal to GQD under a toll processing agreement; the metal feeds Magnequench's NdFeB powder plant in Korat. Neo's filings do not name the other owners; a June 2026 Lanner report describes GQD as Chinese-owned and part of Ganzhou Qiandong Rare Earth Group. Neo said in August 2026 that Silmet sources its feedstock from a Russian supplier. No capacity, output or start date has been published.
Source: Neo AIF 2025 Lanner Jun-26
# Arnold Magnetic Technologies (Thailand) US-owned
Amata City, Chonburi · SmCo
Arnold's SmCo magnet production, shaping/finishing and rotor/stator assembly site, opened Oct 2024; expansion announced 28 Apr 2026 — Phase I complete, Phase II by mid-2027, doubling Arnold's group magnet capacity and relieving the Swiss (Lupfig) plant. Alloy from Arnold's own RECOMA supply. Thailand context: monazite by-product/stockpile exports (~8,200 t in 2025 per DPIM, mostly to Chinese crackers) with no recorded production; USGS 2024 output revised 13,000 → 2,100 t — the 'top-5 producer' headlines are dead. Bonded/ferrite out of scope: Neo Magnequench Korat (1,800 t/yr bonded powder + 1,000 t/yr bonded magnets), Shin-Etsu/TDK/Proterial Thai ferrite/bonded sites. No sintered NdFeB plant announced in Thailand 2023-26. Myanmar Dy/Tb ore flows to China, not Thailand — what crosses the border is arsenic (Kok/Sai/Mekong).
Source: Arnold Apr-26 USGS TH Al Jazeera
Saudi Arabia 1 organization
# Ma'aden + MP Materials + US DoW refinery JV US-owned
term sheet · no site/date
Binding term sheet 18-19 Nov 2025: Ma'aden ≥51%, MP + US Dept. of War 49% (DoW fully finances the US share). Scope = feedstock processing + SEPARATION of light AND heavy RE oxides; magnet manufacturing 'under discussion', not committed. No location, capacity or timeline; not mentioned in MP's Q2-2026 results — slowest-moving item on the map (▸TBD).
Kazakhstan 1 organization
# Lindian — SARECO, Stepnogorsk, Kazakhstan Australian parent
MREC plant · Carester offtake · SX plant at DFS
SARECO hydrometallurgical plant at Stepnogorsk, built in 2010 by Kazatomprom and Sumitomo; Lindian 100% since 10 Aug 2026. Cracks and leaches monazite concentrate to mixed rare earth carbonate (MREC); recommissioning with first MREC targeted Q4 2026 on about 12,500 t/yr of Kangankunde concentrate. A proposed 8,000 t/yr REO solvent-extraction separation plant alongside is at definitive feasibility stage with Carester and Tetra Tech (due end-2026). Binding 10-year Carester offtake (2 Sep 2026): 70% of the heavy (SEGH) product, capped at 8,750 dry t/yr, plus first refusal on 70% of mixed heavy carbonate, delivered to Caremag at Lacq.
Brazil 7 organizations
# Serra Verde (USA Rare Earth) US-owned
Pela Ema, Goiás · ionic clay
Only operating RE mine in Latin America (since Jan 2024): ionic-clay with all four magnet REs incl. Dy/Tb (→ both chemistries' feed). ~4,000 t/yr TREO by end-2026; Stage 2 under construction → 6,400 t/yr within ~12 months. Combination with USA Rare Earth COMPLETED 4 Sep 2026 (~$2.8B; Serra Verde CEO Thras Moraitis becomes USARE CEO 1 Oct 2026); 15-yr 100% offtake with price floors. Mixed-RE product only — no separation on site. Bench: Aclara (Chile/Brazil clays + Louisiana sep/alloy JV w/ CAP, VAC alliance, ▸2028).
Source: USARE 4-Sep-26 Deal Apr-26
# Aclara — Carina Module (Goiás)
ionic clay · Brazil · heavy-RE carbonate · FS Apr 2026 · license pending · pre-FID
Ionic-clay heavy rare earth project in Nova Roma, Goiás, Brazil, 100% owned; makes on-site mixed rare earth carbonate. Feasibility study (NI 43-101, Hatch), 13 Apr 2026: Proven + Probable reserves 170.8 Mt at 1,745 ppm TREO; 18-yr life; 9.7 Mt/yr dry feed. Life-of-mine average 4,378 t/yr REO in carbonate, incl. 1,191 t NdPr, 156 t Dy, 27 t Tb. Capex US$781M incl. contingency; post-tax NPV8 US$1.66B, IRR 26.9% (Argus ex-China prices). Pilot plant in Aparecida de Goiânia since Apr 2025. Preliminary License (EIA filed May 2025) was still under state review in Aug 2026; approval now expected H2 2026 (was Q1 2026). Early works targeted Q1 2027, first production H2 2028. US DFC up to US$5M development funding (Sep 2025). No FID; no binding offtake.
Source: Carina FS Apr-26 Q2-26 MD&A
# Meteoric Resources — Caldeira Australian parent
Minas Gerais · ionic clay to MREC · DFS Jul 2026 · Lynas deal pending
ASX: MEI; OTCQB: METOF. Ionic clay project in the Poços de Caldas alkaline complex, Minas Gerais. DFS (31 Jul 2026): Probable reserve 151.1 Mt at 3,524 ppm TREO; 6 Mt/yr of ore to an on-site plant shipping about 25,000 t/yr of mixed rare earth carbonate containing 12,500 t/yr TREO, 3,862 t/yr NdPr oxide and 127 t/yr Dy+Tb oxide (life-of-mine averages; study figures; the 88,829 t NdPr in some press reports is the life-of-mine total); initial capex US$498M; post-tax NPV8 US$847M at July 2026 spot. Pilot plant made 213.8 kg of carbonate (Jan-May 2026). Preliminary License Dec 2025; Installation License due for consideration in the Dec quarter 2026; no FID or production date. Offtake memoranda (POSCO, Neo, Ucore) and EXIM and EFA finance letters are non-binding. 1 Oct 2026: binding scheme deed for Lynas Rare Earths to acquire Meteoric (0.0207 LYC shares per share), targeted for early Mar 2027, with a Lynas loan of up to A$110M.
Source: Caldeira DFS (ASX, 31 Jul 2026) Lynas scheme presentation (1 Oct 2026) June 2026 quarterly
# Brazilian Rare Earths — Camaçari
Bahia · Monte Alto to Camaçari refinery · first output target 2031
Planned chain: Monte Alto mine (Rocha da Rocha Province, Bahia) feeding a refinery at the Camaçari Petrochemical Complex, about 260 km away. Scoping study update, Aug 18, 2026 (about ±40%; replaces the withdrawn Aug 13 case): about 360,000 t/y ore sorter feed, 9-year mine life, life-of-mine averages of 5,661 t/y NdPr oxide and 2,233 t/y heavy rare earth concentrate (study figures); capex about US$969M, no development funding secured; FID targeted mid-2029, first production 2031. Stage I beneficiation pilot commissioned at Camaçari Sep 8, 2026; Stage II hydrometallurgical pilot targeted for Q2 2027 with R$6.4M co-funding. Carester holds a binding 10-year offtake (Oct 2025) for concentrate carrying up to 150 t/y of contained DyTb.
Source: BRE scoping study update 18-Aug-26 GlobeNewswire 8-Sep-26 (pilot plant)
# Viridis Mining and Minerals — Colossus
Minas Gerais · ionic clay MREC · DFS Aug 2026 · target 2028
Colossus ionic clay project, Poços de Caldas, Minas Gerais. DFS (Aug 20, 2026): 5 Mtpa ore nameplate, 25-year production target averaging 4,446 t/y rare earth oxide in mixed rare earth carbonate (MREC), incl. 2,967 t/y magnet rare earth oxides (study figures); initial capex US$449M; Ore Reserve 200.1 Mt at 2,894 ppm TREO. No FID; financial close targeted Q4 2026, first production 2028. Up to US$120M strategic equity announced Aug 20, 2026. BNDES approved a R$77.5M (about US$15M) facility for the demonstration center on Sep 9, 2026 (documentation in progress); stated remaining senior debt need about US$280M. Solvay offtake is a non-binding LOI. Installation License applied for, not granted. Demonstration plant (100 kg/h ore) made first MREC in May 2026.
Source: Viridis DFS 20-Aug-26 Viridis BNDES approval 9-Sep-26
# Rainbow Rare Earths — Uberaba (with Mosaic)
phosphogypsum · JV with Mosaic planned · PFS started
Phosphogypsum stacks at Mosaic's Uberaba phosphate operation, Minas Gerais. Joint development agreement with Mosaic; intended JV 51% Mosaic, 49% Rainbow, subject to final terms. March 2026 economic assessment: 2.7 Mt/yr, about 1,900 t/yr NdPr oxide, capex US$279M, first production 2030. No resource estimate yet. PFS started September 2026, due H2 2027.
# Brazilian Critical Minerals — Ema, Amazonas Australian parent
ionic clay · in-situ recovery to MREC · BFS
ASX: BCM. Ionic clay deposit near Apuí, Amazonas. Resource 1,071 Mt at 732 ppm TREO (392 Mt Indicated, 681 Mt Inferred), MREO about 25% (190 ppm); no reserve. Bankable feasibility study (30 Jun 2026): in-situ recovery with magnesium sulfate and an on-site plant making about 10,500 t/yr of mixed rare earth carbonate (about 52% TREO) containing about 5,500 t/yr TREO and 1,900 t/yr MREO; Stage 1 capex US$74M, Stage 2 US$27M; NdPr priced at US$108/kg. No first-production date. Trial mining license and state environmental approval under review. No binding offtake (100% uncommitted); Altris appointed Aug 2026 to lead about 24 weeks of front-end engineering.
Source: Ema Rare Earth Project - Bankable Feasib Brazilian Critical Minerals confirms mea FEED appointment, 31 Aug 2026
Chile 1 organization
# Aclara — Penco Module (Biobío)
ionic clay · Chile · heavy-RE carbonate · RCA Jun 2026 · FS due Q4 2026 · pre-FID
Ionic-clay heavy rare earth project near Penco, Biobío, Chile, held by REE Uno (CAP S.A. 20%). M&I resource 27.5 Mt at 2,275 ppm TREO. First EIA withdrawn Mar 2022 over flora and fauna issues; new EIA filed Apr 2023; revised EIA filed Jun 2024. Biobío Environmental Assessment Commission approved it unanimously on 8 Jun 2026 after 3 addenda, 700+ technical and 1,000+ citizen observations. The RCA was issued 22 Jun 2026 and covers the first six years of mine life; a second EIA is being prepared. Design: up to 320 t/h of clay to super-pure carbonate, recycled water, no tailings storage. Feasibility study targeted Q4 2026; latest technical report is the 2021 PEA (774 t/yr REO). Commissioning targeted mid-2028. No FID.
Source: EIA approval Jun-26 Q2-26 MD&A
South Africa 3 organizations
# Steenkampskraal Monazite Mine (SMM)
Western Cape · ~14.5% TREO · Th-rich · Phase-1 concentrator u/c
Historic Anglo American thorium/monazite mine; world-leading grade (~14.4-14.5% TREO ROM; ~2.1% Th). Control now runs through Bora Mining Investments (strategic stake Feb 2024; Prof Enock Mathebula exec chair; Graham Soden CEO) — Bora is also the mining contractor. NNR authorization to exit care & maintenance (Sep 2024) + contractor engagement (Jun 2025). IDC-funded Phase-1 concentrator in four tranches from Sep 2025 (amount undisclosed); GROUNDBREAKING 22 Apr 2026; design 13,400 t/yr monazite concentrate at >50% TREO (~6,700 t TREO). Company targets: commissioning Aug 2026, first revenue before end-2026 — NO commissioning/first-concentrate announcement found as of 10 Sep 2026 after prior slips (Q4-2024 → Q2-2026 → Aug-2026). Downstream 'six-phase' plan (Ce/La removal, magneto-electrochemical separation w/ 'Remedy Group', fluorination, metallization) is unfunded and rests on obscure counterparties; Mintek lab-scale mixed-RE product (Jun 2026) is cracking, not separation. Thorium → Thor Medical (Pb-212 therapy) MoU; Necsa. Offtakes claimed but unnamed. Product for the magnet chain = Th-bearing monazite concentrate needing a foreign cracker. SA produces zero RE today (USGS MCS 2026; reserves 860 kt). Bench (pre-FID): Rainbow Phalaborwa (phosphogypsum, ~1,865 t NdPr + ~80 t Dy/Tb via SX, US$326M, DFS slipping, DFC/TechMet US$50M undrawn, first production 2028-30); Frontier Zandkopsdrift (revived 2026: Carester tech + 7-yr offtake, IDC US$20M, ~3,038 t NdPr + 114 t Dy + 25 t Tb, DFS H1-2027, production ~2030 with separation at Lacq); Galileo/Glenover stalled.
Source: MW groundbreaking MW 6-phase REEx Jun-26 Rainbow Frontier
# Frontier Rare Earths — Zandkopsdrift, Northern Cape
carbonatite · on-site NdPr SX · Carester heavy offtake
Private, Luxembourg-based; listed on the TSX 2010-2015. Mining right and environmental authorization held. Reserve 41.12 Mt at 1.92% TREO (NI 43-101, 2014). Updated study (2022-25): 3,038 t/yr separated NdPr oxide on site plus 114 t/yr Dy and 25 t/yr Tb oxide and manganese sulfate; Phase 1 capex US$628M excluding contingency. DFS started (Dorfner Anzaplan), due H1 2027; first production 2030 (earlier 2028). South Africa's IDC invested US$20M of equity (announced Feb 2026). Carester supplies solvent-extraction technology and has a 7-year agreement to buy the mixed heavy carbonate for separation at Lacq, binding status not stated. KOMIR (Korea) holds 10% of the project. EU strategic project.
Source: Signing of strategic technology supply a Same release via PR Newswire
# Rainbow Rare Earths — Phalaborwa
phosphogypsum stacks · on-site SX to NdPr oxide
LSE: RBW. Recovering rare earths from phosphogypsum stacks left by phosphate processing at Phalaborwa (Rainbow 85%, Bosveld Phosphates 15%). Resource 35.0 Mt at 0.44% TREO, NdPr 29% of TREO (Sep 2024). Plan: 2.2 Mt/yr of gypsum through leach and ion exchange, then on-site solvent extraction to about 1,800 t/yr separated NdPr oxide plus a heavy carbonate (about 60 t/yr Dy, 20 t/yr Tb); separation route changed from K-Tech ion chromatography to solvent extraction in July 2026. Last capex US$326.1M (Dec 2024 interim study, before that change). PFS due Q4 2026, DFS H1 2027; first production 2028 (earlier 2026). DFC's US$50M is a conditional option via TechMet, not closed. Non-binding Neo MoU (Sep 2026) for 40% of NdPr oxide and 65% of the heavy product.
Source: Interim Results for six months to 31 Dec Subscription raises £11.1m from investor
Malawi 2 organizations
# Lindian — Kangankunde, Malawi Australian parent
monazite concentrate · commissioning Q4 2026
ASX: LIN. Monazite carbonatite. Resource 261 Mt at 2.14% TREO (0.5% cut-off, NdPr 20.3% of TREO); probable reserve 23.7 Mt at 2.9% TREO. Stage 1 concentrator (about 55% TREO concentrate; study figure 15,323 t/yr, plant nameplate 20,000 t/yr) under construction: mining began July 2026, commissioning targeted Oct-Nov 2026, first production Q4 2026. Binding Iluka offtake of 6,000 t/yr for about 15 years to Eneabba, plus a US$20M Iluka loan. About 12,500 t/yr of concentrate is to go to its own SARECO plant in Kazakhstan.
Source: May 2026 update Iluka offtake Project page
# Mkango — Songwe Hill, Malawi Canadian parent
carbonatite · on-site MREC · no mining license yet
TSX-V and AIM: MKA (renamed Mkango Magnetic Materials, Sep 2026). Songwe and the Pulawy plant sit in Mkango Rare Earths Limited, which is merging with the Nasdaq SPAC Crown PropTech Acquisitions (pending since a Q4 2025 target). Updated feasibility study (Mar 2026): about 1.0 Mt/yr ore to a mixed rare earth carbonate of 55% TREO made on site; first five years 5,954 t/yr TREO containing 1,953 t/yr NdPr oxide; capex US$325.5M. Reserve 18.1 Mt at 1.16% TREO. No large-scale mining license: the application is pending and Malawi's mining minister said in July 2026 that a moratorium on new licenses is in place. Full production shown for 2029 (earlier targets 2018, 2025). EU strategic project. Only signed project money: US$4.6M DFC development funding. No offtake.
Source: Mkango Announces Results of Updated Feas Updated feasibility study / Pulawy PFS p
Vietnam — foreign-owned plants + domestic upstream 6 organizations
# Dong Pao — Lavreco (Vimico/Vinacomin)
Lai Chau · licensed 2014, never mined · certificate revoked 2026
Dong Pao bastnaesite-barite-fluorite deposit, Tam Duong, Lai Chau. Operator Lavreco (Vimico 55%; TKV holds >98% of Vimico). Investment certificate December 31, 2013; 30-year mining license December 2014. Licensed reserves about 1.1 million t TR2O3; geological reserves >11.3 million t ore (press, no category). Never mined. Lavreco lost VND 7.7 billion in 9M 2025, and the Lai Chau tax authority enforced >VND 137 billion of tax debt (August 2025). Vimico's audited 2025 accounts report the investment certificate was revoked and the project ended in early 2026. A Hanoi pilot approved September 1, 2026 (another company) may process up to 100 t of Dong Pao ore to >95% total REO over 18 months, for testing only.
Source: CafeF Mar-26 VIR Sep-26
# Vietnam Rare Earth JSC (VTRE)
Ha Nam · separation plant · production halted late 2023 · chairman convicted 2025
Private processor founded 2007, based in Ha Nam province, with a separation plant the chairman put at 5,000 t/yr REO (Reuters, September 2023). No mine; in 2023 it said it imported about 1,000 t/yr of refined material from Australia and Russia. Revenue VND 198 billion in 2022. Binding 100 t oxide supply to ASM's Korean Metals Plant (April 2023). Reuters reported production stopped in late 2023. On May 21, 2025 the Hanoi People's Court sentenced the chairman to 16 years for smuggling and accounting violations. Per the case, VTRE bought >3,500 t of Yen Phu ore (18-20% TREO) in 2019-2023, made about 482 t of >95% TREO oxide and exported about 473 t to buyers in Japan, Austria and China. No restart reported as of October 2026.
Source: Tuoi Tre May-25 Soha Oct-26
# Shin-Etsu Magnetic Materials Vietnam (SMMV) Japanese parent
Japanese subsidiary, Dinh Vu IZ, Hai Phong. Est. 2012 (separation/refining — on Shin-Etsu's own IMPORTED oxide/salt feed, not Vietnamese ore; captive to the magnet line) + magnet plant 2016; 2018 expansion made it INTEGRATED separation→press→sinter; Factory 3 Phase 1 completed Mar 2023 → ~2,000-2,200 t/yr. Machining/finishing at other Shin-Etsu SE-Asia sites. Fully built out — no 2025-26 expansion announced. The biggest, most-integrated magnet operation in Vietnam. Recycling: reported as industrial-scale hydrometallurgical magnet recycling at Hai Phong since 2013, inside Shin-Etsu's own separation circuits, taking both its own manufacturing scrap and used magnets returned from customers. Volumes undisclosed; the 2,200 t/yr figure in circulation is magnet capacity, not recycling.
# Baotou INST Magnetic (Vietnam) Chinese parent
Wholly owned · ChiNext 301622 · magnet assemblies · northern Vietnam
INST Magnetic New Materials Vietnam Co., Ltd., wholly owned (via Hong Kong) by Baotou INST Magnetic New Materials (SZSE ChiNext: 301622), a maker of NdFeB magnet application devices and assemblies for consumer electronics. The parent buys sintered NdFeB blanks from Chinese producers; Apple-related sales were 67.42% of revenue in its prospectus. Reuters (August 2023): local approval June 2023, a leased plant in northern Vietnam, a few million dollars initially, assembly lines moved at customers' request, possible second phase. In April 2025 INST said the Vietnam factory was in overseas customers' supply chains and operating normally; its H1 2025 report cites new buildings and fit-out there. No capacity published. Parent 2025 revenue CNY 1.44 billion.
Source: INST IR Apr-25 Reuters Aug-23
# Star Group — SGI Vina Korean parent
Korean subsidiary — SGI Vina Co., $80M plant in Bac Chu Lai Industrial Park, Quang Nam. OPERATING since ~Feb 2025; 4,000 t/yr sintered NdFeB; 1,050 jobs; inputs PrNd/DyFe/TbH (non-China). Serves Hyundai/Mercedes/Stellantis. LinkedIn-promoted.
Source: TheInvestor
# LS Eco Energy + Lynas (Vietnam) Korean parent
LSCV · GSHM · Sm metal ~240 t/yr · trial Dec 2026
LS Eco Energy (KOSPI: 229640; LS Cable & System subsidiary) is adding rare-earth metal making inside its LSCV cable plant in Ho Chi Minh City (brownfield). First line: samarium metal, ~240 t/yr design capacity (company estimate: feed for ~700-1,000 t/yr SmCo magnets); trial production targeted Dec 2026; no commercial-operation date. Operator: new company Global SHRE Metal (GSHM; charter capital KRW 20bn), 58% LS Eco Energy (KRW 11.6bn cash, board 30 Sep 2026). NdPr, Dy and Tb metal planned later; no capacity or dates. Feed: Lynas oxide under a Mar 2026 framework; the definitive processing agreement is not yet announced. Binding 27 Jul 2026: KRW 30bn / ~A$29M zero-coupon convertibles each way (5-yr). Metal is intended for LS Cable's magnet plants (Korea, Virginia). No Vietnamese mining or separation.
Source: SEDaily Sep-26 Lynas/ASX Apr-26
What has changed
Corrections and revisions are recorded here with the date they were made, rather than folded silently into the data. If you cited an earlier version of this atlas, this is where you find out whether it moved.
- On 16 October the Energy Fuels row stopped naming an offtake customer for e-VAC's Sumter plant, because Energy Fuels' 8-K of 13 October says the offtake agreement is in arbitration without naming the customer. The row and the Energy Fuels and VAC profiles now record the agreement as terminated by General Motors: GM's quarterly report for the period ended 30 September 2026 says it terminated a strategic sourcing agreement with an EV traction-motor component supplier that has since brought an arbitration, and Reuters reported on 23 October 2026, citing a person with knowledge of the matter, that GM canceled its supply deal with the Sumter plant. Neither filing names the other party.
- The atlas now covers sintered rare-earth magnets only: sintered NdFeB and SmCo. Bonded and hot-deformed magnets join ferrite, AlNiCo and iron nitride outside its scope. A company that also makes them is drawn only for its sintered work, and a company that makes only those keeps its row and profile but has no bar, with a note on what it does instead. Three rows are now shown that way: Daido Steel and Daido Electronics, which make hot-deformed and bonded NdFeB; Remloy in Bitterfeld, which makes alloy powder for bonded and hot-deformed magnet makers; and GQD Special Material in Thailand, whose metal goes to Magnequench's bonded-powder plant. Bunting's DuBois, Pennsylvania site, which presses bonded magnets, is now drawn for its assemblies of bought-in sintered magnets, and Chengdu Galaxy for its SmCo magnets.
- Every company on the atlas now has a profile. Sixteen were added: Midwest Advanced Materials, India's rare earth permanent magnet (REPM) scheme, Attero Recycling, Lohum, Momentum Technologies, Critical Materials Recycling, REEcycle, the Korea Zinc and Alta Resource venture, TDA Magnetics, Peabody's Rawhide pilot, Kolektor, XMag, GQD Special Material, Dong Pao, Vietnam Rare Earth JSC and Baotou INST's Vietnam plant. Every one of these rows was re-checked and its notes rewritten.
- Moved earlier or into operation: Midwest Advanced Materials' Hyderabad plant changes from planned for 2027 to under construction, with commissioning reported for 2026; Momentum Technologies' Carrollton, Texas line changes from under construction to operating, as the company reported on October 1, 2026; REEcycle's first commercial plant is now dated 2027.
- Moved later or off operating: Vietnam Rare Earth JSC's separation plant changes from operating to planned with no date, since production stopped in late 2023 and no restart has been reported; the REPM scheme's plants are now undated, because 20 technical bids were opened on August 13, 2026 and no award has been made; Lohum's plant is now drawn from metal to machined magnet, as approved in March 2026, and its recycling line is undated.
- Corrections to what the rows show: Attero's separation bar is removed, since no source describes a separated oxide; Baotou INST's Vietnam plant is wholly owned and does assembly, so it changes from a joint venture spanning machining to an operating assembly site; Peabody's Rawhide pilot is drawn through separation, since it is planned to make a mixed concentrate.
- Four rows are renamed: the REPM scheme awardees row becomes India REPM scheme (6,000 t/yr tender); Critical Materials Recycling (TdVib) becomes Critical Materials Recycling — Boone, IA; Peabody's row becomes Peabody — Rawhide rare earth pilot (Gillette, WY), since the other organizations DOE selected are separate projects; and Ningbo XMag becomes XMag (X-Magnet) — Fenghua, Ningbo, now drawn as a machining site only and listed as a watch row, because the company publishes machining equipment and no magnet material or tonnage.
- Evolution Metals & Technologies at Pohang now records the arrival on October 15 of 13 of 57 containers of ULVAC sintered-magnet equipment, and that the release of that date gives no commissioning date for its 10,000 t/yr target, where earlier releases said the end of 2026; filed capacity remains 660 t/yr.
- Twenty-six companies have their first profiles: Meteoric Resources, Ionic Mineral Technologies, Iluka, Arafura, Aclara, Cyclic Materials, HyProMag, Carester, IREL, Ionic Rare Earths, JS Link, LS Eco Energy, the Saskatchewan Research Council, Geomega, Torngat, REEtec, Remloy, MagREEsource, Brazilian Rare Earths, Viridis, Hastings, the POSCO International and ReElement venture, Jiangsu Huahong, Zhongxi Tianma, DOWA and the Daikin, Shin-Etsu, Hitachi and Tokyo Eco Recycle consortium. Thirteen second sites now link to their company's profile, among them Lynas' Malaysian plant, MP Materials' Northlake and Ma'aden ventures, USA Rare Earth's Blacksburg and Serra Verde, both Shin-Etsu Vietnam and Fukui, Arnold's Lupfig and Thailand sites, and both Star Group rows. Every one of these rows was re-checked and its notes rewritten.
- Meteoric Resources' Caldeira row records Lynas' binding scheme to acquire the company, signed 1 October 2026, cites the definitive feasibility study directly, and changes to NdFeB and SmCo, since the planned carbonate carries samarium. Ionic Mineral Technologies, Silicon Ridge, Utah, is added as a watch row.
- Status corrections. Moved later: Torngat to planned for 2030, Brazilian Rare Earths to planned for 2031 (2027 was a pilot stage), Hastings to under construction for 2027, and DOWA to planned with no date, since no source gives 2027. Construction that is not happening: REEtec moves from operating to under construction for 2026, as only its demonstration plant is confirmed running; the POSCO International and ReElement separation venture moves from under construction to planned for 2028, with no site chosen; Cyclic Materials' McBee plant and JS Link's Malaysian plant move from under construction to planned for 2028. LS Eco Energy's Vietnam plant is split into samarium metal under construction for 2026 and NdPr, dysprosium and terbium metal planned with no date.
- Other row changes: IREL's operating samarium-cobalt magnet plant at Visakhapatnam is added; Aclara's Carina and Penco become mining through separation, and Project Dynamo splits into separation and metallization to alloying, planned for 2028; Cyclic Materials' three sites change to NdFeB only, and Mesa loses a separation bar it does not have; MagREEsource loses a metallization step its hydrogen route skips. Four rows are renamed: JS Link — Kuantan, Malaysia; LS Cable & System — Korea magnet plant; Remloy — Bitterfeld (owned by Mkango since August 2026); and Zhongxi Tianma — Liangshan, Shandong, which has no link to China Rare Earth Group.
- GEM Co. is removed. Its 2025 annual report and 2026 half-year report disclose no rare earth site, capacity, output or revenue, and it sold its e-waste unit in November 2025, so the row's operating bars had no current source.
- The Energy Fuels row no longer names an offtake customer for e-VAC's Sumter plant. Energy Fuels' 8-K of 13 October 2026 says the plant's offtake agreement is in arbitration and does not name the customer.
- Mkango's profile now says that Mkango Resources itself bought Remloy, completing in August 2026, so Remloy sits directly under Mkango while HyProMag sits under Maginito. It had named the buyer as "Mkango GmbH", which the company's completion release does not.
- A Filters card replaces the row of buttons above the chart. Its sections are maturity, status, chemistry, process steps performed, equity status and regulatory compliance, and every filter now applies to the map as well as the chart. The regulatory filters are new: DFARS 252.225-7052 before and from 1 January 2027, and ITAR. They are read from plant locations and public statements, are not compliance determinations, and each row in the chart links the statements behind them.
- Adams Magnetic Products' profile said no DFARS statement had been found; its website says it supplies DFARS-compliant materials, and the profile now quotes that, while noting that no public statement names its magnet suppliers or their countries.
- The producer directory below the chart, and the country list on phones, now follow the Filters card too: once any filter is changed from the default view, they show only the organizations that match, with a note saying so and a button that clears the filters. With the default view, the directory still lists every organization.
- Vacuumschmelze's audited accounts are public for the first time, as exhibits to Energy Fuels' 8-K of 13 October 2026: group revenue of EUR 325.1 million in 2025 and EUR 353.5 million in 2024, EUR 211.5 million in the first half of 2026, and net losses in each period. VAC's profile replaces its "no current revenue figure" note with them, and gives the Permanent Magnets division separately: about EUR 79 million in 2025 and EUR 82.4 million in the first half of 2026.
- The same filing gives site-level capacities for VAC's European plants for the first time, as the acquirer's figures: permanent magnets at Hanau 1,000 t/yr and at Ulvila 500 tons, with Hanau's other steps and Horna Streda by process. VAC's and Neorem's profiles carry them, and their entries saying no European tonnage was published are replaced by what the sources now disagree about: the figures against the 2021 EU-wide aggregate, and Ulvila's share of VAC revenue against Neorem's own filed accounts.
- The 8-K discloses that e-VAC's Sumter plant was constructed in connection with an offtake agreement whose customer gave notice of termination on 1 July 2026, and that e-VAC began arbitration on 7 July 2026. The customer is not named. It is recorded as a point to check on the VAC and Energy Fuels profiles, along with the first amendment to the merger agreement of 13 October and the two Energy Fuels statements on when the acquisition will close, which disagree.
- REalloys has its first profile, built from its SEC filings and from the US Army, the Saskatchewan Research Council and the Saskatchewan legislature's committee transcripts, and a new row for its only operating site: PMT Critical Metals in Euclid, Ohio, which makes rare earth metals and alloys at pilot scale and is the source of all of the company's current revenue.
- REalloys' Tooele Army Depot row now records the June 2026 award as what it is, a conditional selection to negotiate a lease for dysprosium and terbium processing, with no signed lease announced by October 14; its status moves from under construction to planned for 2028. The earlier statements that the metallization plant had moved to Tooele from Ohio and was funded from the June placement are not supported by REalloys' own releases and have been removed.
- The Hoidas Lake row is corrected: a December 2024 S-K 1300 technical report filed with the SEC includes a mineral resource estimate. There is still no reserve or economic study, and no field work has been announced in 2026.
- The SRC Rare Earth Processing Facility row is corrected: the 10 t a month of NdPr metal was a target, and SRC told the Saskatchewan legislature in December 2025 that it had made about 20 t in total by toll-smelting; funding is now the C$216.5M capital budget of April 2026; and REalloys' share of the output is given as both its own figure, 80%, and SRC's, 70% rising to 80%.
- The Korean Metals Plant at Ochang, South Korea, owned by Energy Fuels since its purchase of Australian Strategic Materials in August 2026, has its first profile, with the customer orders ASM published: binding purchase orders from Vacuumschmelze for 7.2 t of NdFeB alloy, deliveries to Noveon Magnetics under a 100 t agreement, a 19 t NdPr metal order to Magnequench, and alloy samples to USA Rare Earth.
- Energy Fuels' profile now states eVAC Sumter's 2,000 t/yr as magnet block, with the published steps to 4,000 and 12,000 t/yr, notes that no Sumter production tonnage has been published, and sources the planned American Metals Plant's 2,000 to 4,000 t/yr to research commissioned by ASM. White Mesa's Phase II heavy rare earth targets were added. VAC's profile records its alloy purchases from the Korean Metals Plant.
- Six companies were added, each with a profile. LKAB's demonstration plant in Luleå, Sweden, which is being built to recover phosphorus and a mixed rare earth concentrate from apatite in iron-ore tailings, is drawn under construction. Chemours's Offerman, Georgia mineral sands operation, which sells monazite as a co-product, is drawn as operating mining. Rare Earth Salts in Beatrice, Nebraska, which separates rare earths electrochemically at a plant it calls a commercial demonstration plant, is drawn under consideration because it publishes no capacity. Brazilian Critical Minerals' Ema ionic clay project in Amazonas, Search Minerals' Deep Fox and Foxtrot deposits in Labrador, and N.A.N. MagneTech's planned magnet plant at Naidupeta, India are drawn under consideration.
- Ramaco's Brook Mine row was rewritten and given its first profile. Since February 2026 Ramaco plans to mine and make a mixed rare earth carbonate on site for other companies to refine, so the bar now runs from mining through separation, the convention for on-site carbonate, and its date moved from 2029 to 2031, the year Ramaco's second-quarter release gives for initial production. The notes now state the company's scope, schedule, resource and financing from its own filings.
- Tronox moved from early to mature. Tronox and Chemours both sell monazite from mineral sands as a co-product, and both are now rated on that monazite rather than on a plan to process it; Tronox's planned carbonate plant is still drawn as under consideration, with no investment decision.
- Orano + CEA Liten's magnet pilot line in Grenoble was added in France, with a profile: short-loop recycling of end-of-life magnets back to alloy, strip-cast, pressed and sintered, at about 4 t/yr, making grade N45UH reproducibly as of September 2026. It is drawn as under consideration because Orano says it will license the process to an industrial partner rather than build a plant, and no partner has been named.
- The scope statement changed. It said "sintered NdFeB and SmCo only" and that bonded makers were out of scope, while the map already carried bonded NdFeB makers such as Bunting's DuBois plant. It now states the rule the rows follow: rare-earth magnets only - sintered, bonded and hot-deformed NdFeB, and SmCo - with ferrite, AlNiCo and iron-nitride magnets out of scope, and a company that also makes them shown only for its rare-earth work. No row was added or removed by this change.
- Shin-Etsu Chemical. The capacity table previously carried no figure for magnet production in Japan, and the company's own 2,200 t/yr Vietnam figure of June 2017 was the only magnet capacity on the page. It now also records that Benchmark Mineral Intelligence reported on 13 January 2026 that Shin-Etsu operates around 3,400 t/yr of sintered magnet capacity across Japan and Vietnam combined. The figure is press reporting rather than company disclosure and is not broken down by country; no figure for Japan alone has been published, the 2017 Vietnam figure is unchanged, and a conflict records why the two must not be subtracted.
- Proterial. The capacity table previously carried no company-wide magnet figure, the only plant-level number being a trade report's more than 2,000 metric tons for the China plant. It now also records that Benchmark Mineral Intelligence reported on 13 January 2026 that Proterial produces 2,500–3,000 t/yr of sintered NdFeB for traction motors. The source names no country and no plant and covers traction-motor products only, so the row carries no geography, and the conflicts entry records that it overlaps the China figure by an unknown amount.
- Vulcan Elements + ReElement. This page described the November 2025 Office of Strategic Capital arrangement as “$1.4 billion in Office of Strategic Capital conditional commitments.” That overstates the federal share. The OSC announcement of 3 November 2025 describes two loans totalling $700 million — $620 million to Vulcan Elements and $80 million to ReElement — matched by private capital to reach the $1.4 billion partnership figure. The conflicts entries now state the split and the matching, and record that the Department of War also receives warrants in ReElement. The $620 million and $80 million figures were already correct elsewhere on the page; the error was in one summarizing sentence.
- Eleven organizations were added, each with a profile: Greenland Mines (Sarfartoq, Greenland), AML (Advanced Magnet Lab, Melbourne, Florida), Tronox (Western Australia), Mkango (Songwe Hill in Malawi, and a planned separation plant at Pulawy in Poland), NioCorp (Elk Creek, Nebraska), Australian Rare Earths (Koppamurra, South Australia), Mont Royal (Ashram, Quebec), Frontier Rare Earths (Zandkopsdrift, South Africa), Rainbow Rare Earths (Phalaborwa in South Africa, and Uberaba in Brazil with Mosaic), Defense Metals (Wicheeda, British Columbia) and Leading Edge Materials (Norra Kärr, Sweden). Poland and Sweden are new countries on the map. All but AML are early stage, so they appear on the map when searched for or when Early stage is ticked.
- Neo Performance Materials: its profile now records the September 2026 sale of the Sarfartoq project to Greenland Mines and the offtake rights Neo kept, which Greenland Mines describes as non-binding.
- Three rows showed a raw parent code ("ca" or "de") where a label belonged. They now read Canadian parent and German parent.
- Baotou Tianhe. The capacity table previously stated 12,500 tons/year of sintered NdFeB design capacity as operating, sourced to the company website. It now carries the 9,300 t/yr figure stated in the company's FY2024 and FY2025 annual reports, with a separate 12,300 t/yr target announced for end-2026, and records the website figure as a conflict rather than retiring it. The Inner Mongolia base moved from announced to under construction and gained a 20,000 t/yr preliminary planned capacity. FY2025 and H1 2026 results replaced the FY2024 revenue line. Two further conflicts were recorded, covering the blank-versus-finished output bases and the unspecified export-license instrument.
- AT&M. The revenue note previously reported FY2024 group revenue of RMB 7.573 billion across four business divisions. It now reports FY2025 and H1 2026 results across three divisions, which is what the company reports, and records rare-earth permanent magnet unit revenue of RMB 1.961 billion for the first time. The conflict stating that no magnet-segment revenue is disclosed was replaced by one recording the scope limits of that business-unit figure.
- Innuovo. The revenue note gained H1 2026 results and the magnetics segment line. The conflict stating that no public source breaks out magnetics revenue was replaced: the FY2025 annual report discloses NdFeB revenue of RMB 1.960 billion and the H1 2026 interim a magnetics segment of RMB 1.098 billion. The expansion-timeline conflict was refreshed for a third repetition of the “second half of the year” language, and the ownership conflict narrowed to the still-unpublished shareholding percentage.
- Seven organizations were added, each with a profile: Evolution Metals & Technologies (Pohang, South Korea), American Rare Earths (Halleck Creek, Wyoming), Solcoa (Alameda, California), Gadolyn (Kingston, Tennessee, drawn as unverified), Radify Metals (Campbell, California), and Lindian Resources at two sites, Kangankunde in Malawi and the SARECO plant at Stepnogorsk in Kazakhstan. Malawi and Kazakhstan are new countries on the map.
- Phoenix Tailings: this atlas said that independent reporting in August 2026 put the company's actual output at roughly 200 kg a year. That misread the source. In the Associated Press report the 200 kg figure follows a passage about samarium and reads as the company's samarium capacity, not its total output; no figure for total output has been published. The note now says so.
- Phoenix Tailings: its Recycling bar changed from operating to under consideration. Recycled magnets and disk drives are named only as a possible feed, with no plant or volume published. A Separation bar was added as under consideration for 2028, for the Freedom Facility, whose US$500 million Office of Strategic Capital loan is a conditional commitment for a site not yet chosen.
- Every organization now carries a maturity stage (mature, ramping, developing or early stage) with a one-sentence reason, and a listing status. Early-stage organizations are hidden from the map and chart by default and appear when searched for by name or ticker; the producer directory below still lists every organization.
- The chart has an eighth column, Recycling, at the far right of the stage row, and 41 of the 107 organizations now carry a bar in it. A solid bar means the organization recovers rare earth from material it did not produce — end-of-life magnets, purchased scrap, e-waste, tailings. A hatched bar means it reprocesses only its own production scrap. Seventeen organizations were added to the directory in the same pass, most of them recyclers that the atlas had no row for: Phoenix Tailings, HyProMag USA, HyProMag Ltd, HyProMag GmbH, Heraeus Remloy, MagREEsource, Geomega Resources, Momentum Technologies, Critical Materials Recycling, REEcycle, Jiangsu Huahong Technology, China Rare Earth Tianma, GEM Co., the Daikin air-conditioner compressor consortium, DOWA Eco-System, Attero Recycling and Lohum. An empty cell in the Recycling column means nothing has been published, not that nothing happens: every sintered magnet maker recovers some of its own swarf, and almost none of them publish it, so the column reports disclosure rather than practice. Jiangsu Huahong alone reported 8,794 t of rare earth oxide recovered from scrap in 2025, which exceeds the combined announced capacity of the Western magnet recyclers on this chart.
- Every organization now has its own entry and its own permanent link in the producer directory below, so a single company can be cited without linking the whole atlas.
- Company profiles published for twenty-one organizations, each with founding, ownership, stage-by-stage capacity, reported revenue where a source states one, and the places where public sources disagree. Every figure carries a linked source and a date. Coverage is partial and will extend.
- Company names in the stage chart and in the map's country list now link straight to that company's profile, where one exists. Clicking a country pins its list in place so the names in it can be clicked; clicking the country again, clicking away or pressing Escape releases it.
- MP Materials profile: added a conflicts entry recording that the Dallas Business Journal reported on 18 September 2026 that MP delivered its first magnets to General Motors, while MP's own public comment the same day described itself only as producing magnets for GM's traction motors, and its CFO had placed the start of commercial deliveries at year-end 2026, six days earlier.
- USA Rare Earth profile: added a conflicts entry noting that the 9 September 2026 Blacksburg groundbreaking release's claim of 10,000 t/yr combined capacity with Stillwater exceeds the company's own published Blacksburg and Stillwater figures, which sum to 7,600 t/yr.
- Vulcan Elements + ReElement profile: the Marion Phase 1 capacity row now records that the first commercial-scale production column commissioned there, in August 2026, processes germanium rather than rare earths, and a Minor Source air permit covering all four planned Phase 1 lines was received 19 August 2026; a new conflicts entry distinguishes a separate $25 million Department of War investment (13-14 July 2026) from the existing $1.4 billion in conditional commitments, and another records the North Carolina Utilities Commission's 31 August 2026 procedural waiver for a Benson transmission line against Vulcan's own public description of it as a full construction approval.
- Lynas Rare Earths profile: ownership now records the company's 2 September 2026 confirmation to Reuters that it held takeover discussions earlier in 2026 that did not proceed.
- Ningbo Yunsheng profile: capacity table now also carries H1 2026 finished-magnet sales of 7,190 t, alongside the existing FY2025 annual figure.
- Company profiles published for seven additional organizations: Hangzhou Permanent Magnet Group (HPMG), Ningbo Ninggang, Chengdu Galaxy Magnets, Golden Dragon Rare Earth, Ningbo Konit (Koningda), Magnetfabrik Schramberg (MS-Schramberg), and Magneti Ljubljana. An eighth candidate, "Ningbo XMag," could not be resolved to a single verifiable public identity and was not published; three non-matching candidate entities were found under similar branding.
- Company profiles published for nine additional organizations: China Northern Rare Earth, China Rare Earth Group, Santoku, Electron Energy Corporation (EEC / Permag), Dexter Magnetic Technologies / MCE (part of Permag), Integrated Magnetics, Adams Magnetic Products, Less Common Metals (LCM), and Solvay's La Rochelle rare-earth separation business. Two items surfaced during research remain unresolved as of publication: unconfirmed, single-sourced reports of talks between China Rare Earth Group and Shenghe Resources, and between Solvay and OneIM, over possible ownership or investment changes; neither is treated as confirmed.
- Company profiles published for the three organizations flagged as unwritten in the company-profiles handoff: Baotou Tianhe Magnetics, Advanced Technology & Materials (AT&M), and Zhejiang Innuovo Magnetics. The open identity question on Zhejiang Innuovo's stock code is resolved: Shenzhen 000795 is confirmed correct for Innuovo Technology Co., Ltd., the listed parent; Shanghai 603072, once suggested as a possible alternative, belongs to the unrelated Baotou Tianhe Magnetics and is not connected to Innuovo or Hengdian Group.
- Data revision 1.9: 90 organizations across 22 countries.
More from Peculiar Materials
- How magnets are made: mine to magnet The thirteen processes between an ore body and a finished magnet, with an animated figure for each one.
- The NdFeB grade chart, explained Why remanence and coercivity trade against each other, and what dysprosium cost to buy that trade.
- Landed cost estimator What a sintered NdFeB magnet lands at from five places at once, split into material, processing, tooling, logistics and duty.