Data as of Jul 25, 2026 · Based on 2,718,867 AI responses across 9,511 prompts · See how Parse measures this
MEGNet is a graph network framework for universal machine learning in materials science, developed by the Materials Virtual Lab. It provides pre-trained models for predicting properties of molecules and crystals, such as formation energy and band gap, with low prediction errors.
Parse Score