Hi @NourNashed,
Here are the options I would use:
- Save Columns (Prediction Formula), and as suggested by @txnelson use the Graph Builder to plot the response surface model depending on two factors (use two factors as X and Y variables, Prediction Formula as Color, and click on Contour icon to display the prediction surface response). Example here on Borehole uniform dataset from JMP:

- Save prediction formula column, and use Contour Plots platform (Graph>Contour Plot). Specify as Y your prediction formula and your inputs as X variables, and check the option "Fill Areas". From there, you should have a similar visualisation, with other options in the red triangle (like "Show data points" that I used in the following capture):
I think the difference in your plots are linked to the fact that in one case you're using predicted values (contour plot figure a), and in the other you're using only measured values (figure b).
Hope these options will help you,
Victor GUILLER
"It is not unusual for a well-designed experiment to analyze itself" (Box, Hunter and Hunter)