CederGroupHub/chgnet

Pretrained universal neural network potential for charge-informed atomistic modeling https://chgnet.lbl.gov

72
/ 100
Verified

CHGNet helps materials scientists and researchers quickly predict the properties and behavior of new materials by simulating atomic interactions. It takes in a material's atomic structure and provides highly accurate predictions for properties like energy, forces, stress, and magnetic moments, similar to expensive DFT calculations but much faster. This tool is ideal for accelerating materials discovery and understanding.

368 stars. Used by 3 other packages. Available on PyPI.

Use this if you need to perform accurate, charge-informed atomistic modeling and property prediction for novel materials without the computational cost of traditional Density Functional Theory (DFT).

Not ideal if your research does not involve material science or atomistic simulations, or if you require atomistic modeling with precision exceeding near-DFT accuracy for highly specialized use cases.

materials-science computational-chemistry atomic-simulation materials-discovery solid-state-physics
Maintenance 10 / 25
Adoption 13 / 25
Maturity 25 / 25
Community 24 / 25

How are scores calculated?

Stars

368

Forks

96

Language

Python

License

Last pushed

Feb 19, 2026

Commits (30d)

0

Dependencies

7

Reverse dependents

3

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/CederGroupHub/chgnet"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.