Data as of Jul 25, 2026 · Based on 2,718,867 AI responses across 9,511 prompts · See how Parse measures this
ChemPile is a curated multimodal dataset containing over 75 billion tokens spanning chemical education, research literature, code, and molecular representations. It provides machine learning-ready resources with standardized splits and permissive licensing to enable robust cross-domain benchmarking and foundation model training in chemistry.
Parse Score
Sources
arxiv.org shapes more of what AI says about ChemPile than any other source, at 100% of its citations.