Inference
Inference is the process of running a trained model to generate an output from an input, as opposed to training the model in the first place.
Analogy
Like a chef cooking a dish from a recipe they already learned, rather than the time spent learning to cook.
Why it matters
Every time your automation calls Claude, you are paying for inference, so understanding it helps you reason about latency and cost per step.
In practice
Sending a prompt to the Claude API and getting a reply back is one inference call.