This dataset contains DFT computations on 550 halide perovskite compounds, 90 of which are pure ABX3 compositions and the remaining involve mixing/alloying at the A, B or X sites. Computation folders include geometry optimization using GGA-PBE and HSE06, optical absorption calculations, vacancy defect calculations, and dielectric constant calculations. Machine learning models were trained using this data to accelerate the prediction of stability and optoelectronic properties of a combinatorial perovskite dataset. One publication is currently under review: https://arxiv.org/abs/2109.10798, and two others are in preparation.