Inference

Inference is the process of running a trained AI model to generate an output.

Every time a model answers a question, writes code, analyzes an image or takes an action, computation is being performed during inference.

Training creates the model. Inference is the model being used.