BEHAVIOR-1K is a comprehensive simulation benchmark platform for embodied AI research developed by Stanford. It provides all components needed to train and evaluate AI agents on 1,000 everyday household activities including cleaning, cooking, and organizing. Tasks are selected from real human time-use surveys and preference studies to ensure realistic, human-centered benchmarks for advancing embodied AI systems.