Pandas uses Matplotlib behind the curtains to generate the graphs, so there’s no difference in terms of output — what you get with Pandas, you can get with Matplotlib.
Using Pandas visualization methods can be quicker if your data comes in the form of Pandas objects, like
DataFrame. When you’ve just read in a dataset and want to explore it using data visualization, using Pandas visualization methods will make you work faster. The Pandas visualization ecosystem is quite complex — see a full tutorial here.
Matplotlib is great for advanced graph customization (see the Storytelling Data Visualization and Information Design course) or for creating graphs out of non-Pandas objects.
All in all, feel free to use what you’re accustomed with. One of the reasons we teach both ways is that learners need to know to at least read both — although they can only use one to generate their own graphs.