These datasets were employed in the manuscript: [1] J. Järvi, M. Todorović, and P. Rinke, Efficient modeling of organic adsorbates on oxygen-intercalated graphene on Ir(111), Phys. Rev. B 105, 195304 (2022). DOI: 10.1103/PhysRevB.105.195304 The two datasets feature DFT sampling of adsorption configurations for F4TCNQ and TTF on approximate supported graphene (Gr) using the Bayesian Optimization Structure Search (BOSS) method. The 3D configurational searches were conducted for each molecule over two molecular translations (x, y) and an in-plane rotation (gamma). The search range of on-surface translation (x, y) is [0, 1] in fractional coordinates of the Gr unit cell. The in-plane rotation (gamma) is done in the range [0, 180] deg. Energy (Eads) is the adsorption energy of F4TCNQ or TTF on the approximate supported Gr, computed with DFT. For additional details on search coordinates and DFT calculations please refer to [1]. The 3D models of the adsorption energy surfaces were constructed with 500 energy points for each molecule, the full lists are provided in the datasets. Each calculation is denoted with an ID number (001-500). The raw data for each calculation can be found in the NOMAD repository with the corresponding ID, DOI: 10.17172/NOMAD/2022.04.07-1