You are asking for something practically impossible yet.There are dozens of high quality visualization programs(e.g.Rasmol,Chimera,Maestro etc) but they all require known or modeled input structures i.e cartesian coordinates.Approximate predictions may be obtained for small-intermediate size proteins via homology modelling,energy minimizations and MD with special programs but it is hard to estimate their accuracy and
For protein visualization I would recommend Pymol. There is a free version of the software, however the most updated is payed.
For protein tertiary structure prediction there are a bunch of web servers that would scan for your sequence to predict the 3D structure if they find something similar in the PDB.
Basic and based on homology modelling: Phyre2
More advanced and combining homology modelling and threading: I-Tasser
Protein visualisation may be achieved by using Pymol, Biovia Discovery Studio or Swiss PDB Viewer. Concerning protein tertiary structure prediction, I suggest swiss-model server.