Large datasets are becoming increasingly important in energy research, and Nature Energy has introduced the Resource article format to help disseminate these datasets. The growth of large datasets is driven by big data opportunities and high-throughput techniques, enabling scientific research and AI-based discoveries. Nature Energy now publishes datasets as Resource articles, which are similar to research articles but focus on presenting large datasets, new data platforms, or useful libraries. These articles should explain the resource's utility and include detailed methods and data acquisition. Data, algorithms, and codes should be made publicly available to ensure reusability. Machine-readable data is essential for AI applications, following the FAIR principles (findable, accessible, interoperable, reusable). Resource articles can take various forms, such as databases with analytical tools or generated datasets from existing data. Examples include a database of perovskite solar cell data and a high-resolution meteorological dataset generated using machine learning. These resources help bridge knowledge gaps in energy fields. Nature Energy encourages authors to submit studies that effectively highlight their findings, and editors will work with authors to determine the best format. The journal expects more studies that explore both existing and new scientific areas, advancing energy system understanding and application. Published online: 25 July 2024.Large datasets are becoming increasingly important in energy research, and Nature Energy has introduced the Resource article format to help disseminate these datasets. The growth of large datasets is driven by big data opportunities and high-throughput techniques, enabling scientific research and AI-based discoveries. Nature Energy now publishes datasets as Resource articles, which are similar to research articles but focus on presenting large datasets, new data platforms, or useful libraries. These articles should explain the resource's utility and include detailed methods and data acquisition. Data, algorithms, and codes should be made publicly available to ensure reusability. Machine-readable data is essential for AI applications, following the FAIR principles (findable, accessible, interoperable, reusable). Resource articles can take various forms, such as databases with analytical tools or generated datasets from existing data. Examples include a database of perovskite solar cell data and a high-resolution meteorological dataset generated using machine learning. These resources help bridge knowledge gaps in energy fields. Nature Energy encourages authors to submit studies that effectively highlight their findings, and editors will work with authors to determine the best format. The journal expects more studies that explore both existing and new scientific areas, advancing energy system understanding and application. Published online: 25 July 2024.