As of today, you can explore projects of up to 107MB. The first explore carried out is slower than subsequent iterations rest since there are initialization issues to handle.
As the length of a verbatim is very variable, the exact performance depends on the actual size of the file. If your file is too large, you can experience time outs and other issue with performance. However, there are almost always ways to cut down your amount of data for each individual project. It would be highly unusual to have a text file of more than 5MB which is not categorisable in any way. It usually makes sense to sort your data by some kind of meta-criteria, such as time period.
Once you have smaller data files to work with, it is easy to develop an explorer model on one project and apply the same model to subsequent projects with verbatims from the same source.