Slope stability assessment plays a crucial role for geotechnical engineers as failure could pose a devastation in the hilly regions. This paper presents a performance comparison of three machine learning models including artificial neural network (ANN), random forest (RF) and multiple linear regression analysis (MLRA) for estimating safety factor of circular failure slope. A total 130 slope cases were considered from the past literatures from which the database is subdivided into training and testing with a ratio of 70:30. Distinct geometric and shear strength parameters such as slope height (H), angle (β), internal friction angle (ϕ), cohesion (c), pore pressure ratio (ru) and unit weight (γ) were taken as an input parameter of model. However, factor of safety (FS) denotes the target parameter. The ANN model with 6–8-1 network structure produce the best results with sigmoid as the finest activation function for the precise forecast of FS. For validating the model performance, the proposed model used the testing dataset. The result revealed that the optimized ANN model acquires higher accuracy over RF and MLRA with a coefficient of determination (R2) value of 0.97, 0.86 and 0.82, respectively. Furthermore, the relative importance of variables has been estimated by sensitivity analysis, which ascertains that β, H and ϕ as important parameter for FS prediction. The developed model can be used as a necessary tool for assessing slope stability during primary stage of designing projects along the soil slope.