AI-driven discovery of superior permanent magnets. Our goal is to develop magnets that are powerful, cost-effective, easy to manufacture, and free from rare-earth minerals.
RE-free magnet leaderboard is open: submit your candidate
The leaderboard is live and open for entries.
Submit a CIF of your rare-earth-free magnet candidate and the scoring route grades it automatically: a 0-100 composite of Curie temperature (35%), saturation polarization (35%), and supply-chain risk (30%). No lanthanides; Y allowed. Multiple entries welcome, one CIF per entry.
A few honest limits, stated up front. Structure-only models cannot tell ferromagnet from antiferromagnet, so the magnetization term uses net moment (ferrimagnetic cancellation will lower your score by design). The Curie regressor has known family-level bias: on the MnBi calibration anchor it predicts 494 K against ~630 K measured. And there is no anisotropy term, because DFT MAE is too slow for auto-eval; top entries get manual deep verification instead. Scores rank, they do not certify.
Verifying the MnBi₁₋ₓGeₓ permanent-magnetism claims: both arms complete, falsification criteria answered
This post was first published 2026-08-27 as an interim report with the parent control done and the Ge-substituted runs in flight. Both Ge-substituted arms have since landed; the completed results and the outcomes of the pre-registered falsification criteria are folded into the body below (also posted as comments, preserved verbatim).
GGen ran 890 trials across 89 Co-Fe-Hf stoichiometries in 17.6 min, keeping 839 relaxed structures. 42 of them landed within 150 meV/atom of the convex hull, 41 of which the database had not seen before. The leading generated candidate is CoFe15 in Cmmm (#65), which sits on the hull (known).
GGen ran 1,000 trials across 100 Co-Fe-Sb-W stoichiometries in 26.0 min, keeping 951 relaxed structures. 14 of them landed within 150 meV/atom of the convex hull, 14 of which the database had not seen before. The leading generated candidate is Co3Fe9SbW in Pm (#6), which sits 82 meV/atom above the hull (new composition).
GGen ran 1,000 trials across 100 Bi-Co-Fe-W stoichiometries in 25.2 min, keeping 958 relaxed structures. 7 of them landed within 150 meV/atom of the convex hull, 7 of which the database had not seen before. The leading generated candidate is CoFe13W in Cm (#8), which sits 7 meV/atom above the hull (new composition).
GGen ran 1,000 trials across 100 Co-Fe-Ta stoichiometries in 19.8 min, keeping 970 relaxed structures. 68 of them landed within 150 meV/atom of the convex hull, 66 of which the database had not seen before. The leading generated candidate is CoFe6 in R-3m (#166), which sits 1 meV/atom above the hull (known).
GGen ran 1,000 trials across 100 Co-Fe-W stoichiometries in 20.0 min, keeping 928 relaxed structures. 26 of them landed within 150 meV/atom of the convex hull, 26 of which the database had not seen before. The leading generated candidate is Fe13W2 in C2/m (#12), which sits 7 meV/atom above the hull (new composition).