Term of the Moment

Internet modem


Look Up Another Term


Redirected from: AI execution speed

Definition: inference speed


The time it takes to generate an answer from an AI chatbot. The inference speed is the time between a user asking a question and getting an answer. It is the execution speed that people actually witness. Although considerably more time consuming (days, weeks and months), people never experience the training phases that developed the models that the inference engine uses. See AI inference vs. training.