The sandbox lets you test requests to the BibleGet service endpoint query.bibleget.io as an application would make them. The results you see are as the application would see them, not as the final user would see them. This means that you will see raw JSON, XML, or HTML without any kind of formatting. If you would like to see an example of formatting as the final user might see it (this however depends on how the application processes the raw data and formats it for the user), you can choose "HTML" output and then check "view sample formatting".
Please note that this sandbox page makes use of WordPress transients as a caching mechanism. For now, the BibleGet project does not require developers to use an access key in order to make use of the service. Please do not be indiscriminate in how you use the endpoint. If for example you implement calls to the endpoint from a webpage, please make use of a server side caching mechanism or even client side browser storage to avoid making the same calls over and over again, bogging down the server. You may not realize in fact that on a webpage, not only you but any other internet user or perhaps even an indexing robot from a search indexing tool such as Google could access the page you created, triggering a new request to the BibleGet endpoint. For this reason the more recent updates to the BibleGet engine will do a few checks to see if too many requests are coming from the same source, and if so will start to show a notice or perhaps even deny service for the request.