A 3-dimensional heatmap is fairly esoteric need. There are few situations in which they let one clearly visualize the data (because the surface of your graph will obstruct your view of what's going on in the middle), so not many people use them.
Also, I would recommend against picking a tools based on the input format; you're going to be better off planning to transform your data from the format that's easy for you into the format that your tool wants.
The above two points combine into a piece of bad news: you'll probably have to write a program to do the thing you want.
The good news is that there are tools (very nice ones actually) for making your own data visualizations.
Mathematica is a dream to work in, but it's expensive. (Actually the price has come down over the years, but still...)
I haven't used MatLab, but supposedly it's a fine choice.
And for open-source stuff, Jupyter Notebooks with matplotlib approaches the user experience I miss so much from Mathematica, although you'll have to learn Python.
What I think you're actually going to end up doing is building a 3D scatter-plot, either with your data split into buckets of different hues based on the 'density' value, or with only points who's 'density' exceeds some threshold included in the output. All of the tools I've listed above can do that.