Running an AI model to generate outputs in real time (e.g., answering queries, powering chatbots). Inference is different from training because it is ongoing and usage-based, creating recurring revenue.
« Back to Glossary IndexInference Workload
« Back to Glossary Index