Managing Private Endpoints

A private endpoint is a private IP address within a VCN that you can use to access a specific service in OCI. The service sets up the private endpoint in a subnet of your choice within the VCN. Think of the private endpoint as another VNIC in the VCN. You control access to it similar to any other VNIC, using security rules. The service creates this VNIC and maintains its availability for you. You only need to manage the subnet and its security rules. Learn about About Private Endpoints.

When you create a private endpoint, you receive a fully qualified domain name (FQDN) for it, regardless of whether you select the Allow Usage In On-Demand Mode option. You can use the private endpoint in the following ways:

• Attach it to an endpoint on a dedicated AI cluster to make the cluster (and its hosted models) accessible through the private endpoint.
• If you select Allow Usage In On-Demand Mode during creation or editing, you can also use the private endpoint to access on-demand models offered in the Generative AI service.

To access a model through a private endpoint, you can create a Compute instance in the private subnet allocated for the private endpoint, add your code to the Compute instance, use the FQDN for the private endpoint, and access the models from that Compute instance.

Before creating a private endpoint, perform the Prerequisites for Private Endpoints and set up a virtual cloud network with a private subnet.

By default, a tenancy has a maximum limit of 5 private endpoints. For more Generative AI private endpoints, request an increase in the limit, private-endpoint-count for the Generative AI service.

You can perform the following tasks to create and manage endpoints for custom models:

• Creating a private endpoint
• Listing the private endpoints
• Getting a private endpoint's details
• Updating a private endpoint
• Moving a private endpoint
• Deleting a private endpoint

Learn about private endpoints in OCI.