This study systematically compares probabilistic slope stability analyses performed using three widely used geotechnical engineering software packages: PLAXIS LE V21, GeoStudio 2024 (SLOPE/W module) and Slide2. Probabilistic analysis has emerged as an essential approach for quantifying uncertainties and calculating key metrics such as probability of failure and reliability index, given the critical importance of risk assessment and the inherent uncertainty in soil parameters. This research evaluates the capabilities, accuracy and efficiency of each software package, as well as their respective limitations, by performing identical analyses on three distinct scenarios (homogeneous soil, three-layered soil and pseudo-static conditions) while employing ten common limit equilibrium methods. The results show that, as the complexity of the problem increases, the factor of safety decreases while the probability of failure and discrepancies between the software packages increase. In the homogeneous scenario, the mean factor of safety ranges from 1.35 to 1.55, depending on the method selected, with a failure probability of 8–12%, and inter-software differences of less than 5%. In the layered scenario, the mean factor of safety decreases to 1.30–1.40, with inter-software discrepancies reaching approximately 15%. Under pseudo-static conditions, the mean factor of safety reduces by around 21% (to 1.15), the probability of failure rises to an average of 27%, and the inter-software discrepancies reach 25%. Advanced methods (Morgenstern-Price and Spencer) yield higher safety factors than simple methods (Ordinary/Fellenius). In terms of software performance, Plaxis LE offers the greatest accuracy in complex conditions, GeoStudio provides the most conservative estimates and, thanks to its advanced graphical tools, Slide2 is a suitable option for probabilistic risk assessment.
