MACE
AI Models Apache-2.0 (Free)
About
Equivariant message passing neural networks achieving state-of-the-art accuracy for atomistic simulations. Universal foundation models for molecular dynamics and materials science with DFT-level accuracy and computational efficiency.
Citation
Batatia, I. et al. MACE: Higher Order Equivariant Message Passing Neural Networks for Fast and Accurate Force Fields. NeurIPS (2022). DOI:10.48550/arXiv.2206.07697
Frequently Asked Questions
What is MACE?
MACE is a ai models application available in the Paramus App Store. Equivariant message passing neural networks achieving state-of-the-art accuracy for atomistic simulations. Universal foundation models for molecular dynamics and materials science with DFT-level accuracy and computational efficiency.
Is MACE free to use?
Yes. MACE is distributed under the Apache-2.0 (Free) license and is available at no cost through the Paramus App Store.
How do I install MACE?
MACE is installed through Paramus Chemistry OS, an on-premise Windows platform for computational chemistry. Open the Paramus App Store in your local installation and select MACE for one-click deployment.
What type of application is MACE?
MACE belongs to the “AI Models” category in the Paramus App Store. It runs on Paramus Chemistry OS and can also be accessed through Paramus Cloud for supported workflows.
What platform does MACE run on?
MACE runs on Paramus Chemistry OS, a Windows-based on-premise platform that provides local compute power for demanding simulations. It requires a Paramus OS installation with appropriate hardware resources.
Can MACE be automated or integrated with AI workflows?
Yes. MACE is available as part of the Paramus ecosystem which supports MCP (Model Context Protocol) tools for AI-driven automation. This enables integration with large language models and automated research pipelines.
How should I cite MACE in publications?
The recommended citation for MACE is: Batatia, I. et al. MACE: Higher Order Equivariant Message Passing Neural Networks for Fast and Accurate Force Fields. NeurIPS (2022). DOI:10.48550/arXiv.2206.07697