Offline model:

Phrase not often used, but contrast with online model. Most machine learning models are offline in the sense that training is done once on some separate system and then inference is done in production on the trained model, but no more training is being done on the new data.