I’m assuming the names are somehow related to image filenames, like Julius.jpg and Claudius.jpg. Other schemes are possible. It’s just that unique data needs to be defined somewhere and related to any other data needed. A variation could be storing such data in the meta data of attachment posts which are already automatically created when uploading media. Then all we need is a reference to the attachment, such as its ID or filename base.

It makes little difference now whether script creates 900 posts or a template indirectly references 900 image pairs. But in the future, should anything need to be changed or added to these posts, it will be far easier to alter the one page template than write a new script to alter 900 posts. As far as site maintainability goes, a page template is much superior to automated post creation.