Difference in R and Python forecast results using same time series Algorithm

Question

If I am using Air Passenger data and doing basic time series forecasting using ARIMA and calculate RMSE (in R and Python), then we get RMSE ==== in Python= 25.6 and R=23.5, which means forecasts are different. They are vastly off in certain other scenarios and other models as well. Can anybody help me understand reason behind the same ? Appreciate in advance for any pointers