In the API model, whenever a service needs something, it makes a network call to a known endpoint with a request and gets a response back. The service making the call is frequently called the client, and the other is the server. It should be noted that a service can be (and is frequently) both a client and a server in the context of different interactions.