Kimola models can represent various things at different time points.
When you are creating a new model, a model represents a space where you can push the data that you want to train the Kimola Algorithm.
When you are training the algorithm, after the first record is pushed, it represents a trained Kimola algorithm and its training set. As you update its training set by pushing records into the model or deleting records from the model, the model algorithm also gets updated.
When you are analyzing your text, a model will be just a trained algorithm and will give you the analysis of your text.