concept
Training is when a model learns; inference is when it is used. Every time you send a prompt and get a response, the model is performing inference: computing an output from the patterns it already holds.
During ordinary inference the model’s parameters do not change, so using a model does not, by itself, teach it anything new.
Each run of inference takes computing power, which is one reason fast and efficient models matter.
Example
When an app turns a voice memo into text, it is running inference on a speech model.