atomicarchitects/nequix
[NeurIPS'25 AI4Mat] Nequix: Training a foundation model for materials on a budget and [arXiv'26] Phonon fine-tuning (PFT)
This project helps materials scientists and researchers quickly and accurately predict material properties. It takes atomic structure data as input and produces highly accurate predictions for forces, energies, and phonon properties of materials, using pre-trained Nequix models or models fine-tuned for specific applications. Researchers who need to simulate material behavior for design or analysis would use this.
Available on PyPI.
Use this if you need to perform high-fidelity simulations of material properties, like forces, energies, and especially phonon calculations, on large datasets without the computational cost of traditional Density Functional Theory (DFT) methods.
Not ideal if you require quantum-level accuracy for very niche or exotic material interactions that are not well-represented in existing DFT-derived datasets.
Stars
68
Forks
14
Language
Python
License
MIT
Category
Last pushed
Mar 11, 2026
Commits (30d)
0
Dependencies
14
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/atomicarchitects/nequix"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Compare
Related frameworks
deepmodeling/deepmd-kit
A deep learning package for many-body potential energy representation and molecular dynamics
chemprop/chemprop
Message Passing Neural Networks for Molecule Property Prediction
Acellera/moleculekit
MoleculeKit: Your favorite molecule manipulation kit
mir-group/nequip
NequIP is a code for building E(3)-equivariant interatomic potentials
CederGroupHub/chgnet
Pretrained universal neural network potential for charge-informed atomistic modeling...