Rendering text in a material

there is tutorial for it in 3d max:

When you have more than one material, you just select which one material you change everything else is same.

is there a way for editing which faces gets which materials without using 3d max?

Well whole thread is about doing material that you can manipulate from unreal and apply different material parts to faces.
You need to decide if unreal does heavy work of stitching together material for those boxes, or you do it in 3d application.

IMO you should get blender, find out how to make sub materials there, and do your box proper way. Its just box, perfect example to learn very basic stuff.
If you don’t do it now, next time you will have problem with gun, chair or something else that you cannot find ready to use. Or if you ever find shape you need it will
have multiple errors that need to be fixed for unreal. So do this thing now and learn new stuff.

Your problem would be very trivial if you could make 16 boxes, each for different number. Then there would be no need for complicated materials etc.

Yes but the whole point for me is to make it as flexible as I can so I can change it to a picture later on for example.
I have a student license for 3DS Max, but I want to understand if I can fix this simple uv mapping in Unreal.
This is why I’m asking specifically about how to do it inside the unreal engine.

I’m surprised no one has recommended to try using a SceneCapture2D component. The general idea is that you would create Render Targets for each face of your cube/tile. By default, render targets are 256x256 but you can change that based on your needs. In your level, you would create a SceneCapture2D component and set its output to a Render Target. The Render Target is just a texture, so you would then create a material that uses that render target and apply that to the face of your tile.

Then to put something onto the face of your tile, you just shove it in front of the relevant Scene Capture 2D camera. The advantage here is that you could project anything on the cube face, particles, text, other 3D models, a 3D scene etc. The disadvantage is that it could be a lot of overhead for a lot of tiles.

Here’s a simple example of some spherical eyes floating in space, the arrow represents where the Scene Capture 2D component is.

And here is that capture component mapped to the face of a model:

One of the cool things about capture components is that they have extra options to disable specific render layers from their output so you don’t need to do any masking in your material.