When interpreting ERT data, it is important to evaluate the model fit to the data and assess the uncertainty in the model parameters. Here are a few steps you can follow to evaluate which model is the best to carry out interpretation:

In summary, evaluating the model fit to the data, assessing parameter uncertainty, considering geologic information, and using multiple models can help determine which model is the best to carry out interpretation.

Always remember that there is a level of uncertainty in geophysics. It is called non-uniqueness. It means that more than 1 model can explain your results. You are inverting your data to get the model so it depends on many factors (input data, software, inversion algorithms, iteration numbers and error). So here, the errors are close yet the values of resistivity are different. You can choose setting the same color bar in Res2D and see if they are more similar after that. Also the iterations are different as well.

Let's start with the model's performance and revisit some of the other considerations to keep in mind when selecting a model to solve a problem.

Let's start with the model's performance and revisit some of the other considerations to keep in mind when selecting a model to solve a problem.

for me I see there is no significant differences between the obtained models, and to answer your question about (How to choose optimum model between many models) you have to integrate your models with another tool to achieve integration and see if both techniques provide good matching results for the problem at hand, if this is not available, thus, I can guarantee you that there nothing to do except studying the geological setting of the study area very well especially outcrops at or near your study area.