atomind-ai/mlip-arena

🌟 [NeurIPS '25 Spotlight] Fair and transparent benchmark of machine learning interatomic potentials (MLIPs), beyond basic error metrics https://openreview.net/forum?id=SAT0KPA5UO

55
/ 100
Established

This platform helps materials scientists and computational chemists fairly evaluate different machine learning interatomic potentials (MLIPs). It takes various MLIP models as input and produces insights into their physical soundness and real-world performance, going beyond simple error metrics. The main users are researchers and developers who need to select or improve MLIPs for molecular and materials modeling.

Available on PyPI.

Use this if you need to rigorously benchmark and understand the strengths and weaknesses of different MLIPs for tasks like molecular dynamics simulations, especially when comparing models trained on diverse datasets.

Not ideal if you are looking for a simple, quick comparison based solely on basic error statistics, or if you primarily work with traditional DFT methods without incorporating MLIPs.

materials-modeling computational-chemistry molecular-dynamics MLIP-evaluation potential-energy-surfaces
Maintenance 10 / 25
Adoption 9 / 25
Maturity 25 / 25
Community 11 / 25

How are scores calculated?

Stars

92

Forks

8

Language

Jupyter Notebook

License

Apache-2.0

Last pushed

Feb 23, 2026

Commits (30d)

0

Dependencies

12

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/atomind-ai/mlip-arena"

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