Models are deployed as web services. You can access the services through the management console or APIs.
A batch service performs inference on batch data and automatically stops after data processing is completed.
A batch service processes batch data at a time. A real-time service provides APIs for you to call.