Some Gemini Enterprise Agent Platform service producers require you to connect to their
services through
[Private Service Connect endpoints](https://docs.cloud.google.com/vpc/docs/private-service-connect.md.txt#endpoints).
These services
are listed in the
[Gemini Enterprise Agent Platform access methods](https://docs.cloud.google.com/gemini-enterprise-agent-platform/machine-learning/general/netsec-overview.md.txt#access-methods)
table. They support unidirectional communication
from a service consumer's on-premises, multicloud, and VPC workloads to
Google-managed Gemini Enterprise Agent Platform services. Clients connect to the endpoint
by using internal IP addresses. Private Service Connect performs
network address translation (NAT) to route requests to the service.

Service consumers can use their own internal IP addresses to access these
Gemini Enterprise Agent Platform services without leaving their VPC networks
or using external IP addresses by creating a consumer endpoint.
The endpoint connects to services in another VPC network using a
Private Service Connect forwarding rule.

On the service producer's side of the private connection, there is a
VPC network where your service resources are provisioned. This
network is created exclusively for you and contains only your resources.

The following diagram shows a Vector Search architecture in which the
Vector Search API is enabled and managed in a service project
(`serviceproject`) as part of a [Shared VPC](https://docs.cloud.google.com/vpc/docs/shared-vpc.md.txt)
deployment. The Vector Search Compute Engine resources
are deployed as a Google-managed Infrastructure-as-a-Service (IaaS) in the
service producer's VPC network.

Private Service Connect endpoints are deployed in the service
consumer's VPC network (`hostproject`) for index query, in
addition to
[Private Service Connect endpoints for Google APIs](https://docs.cloud.google.com/gemini-enterprise-agent-platform/machine-learning/general/googleapi-access-methods.md.txt#psc)
for private index creation.

For more information, see
[Private Service Connect endpoints](https://docs.cloud.google.com/vpc/docs/private-service-connect.md.txt#endpoints).

![image](https://docs.cloud.google.com/gemini-enterprise-agent-platform/machine-learning/general/images/vertex-ai-psc-endpoints.png)

Before you configure Private Service Connect endpoints, learn about
access
[considerations](https://docs.cloud.google.com/vpc/docs/about-accessing-vpc-hosted-services-endpoints.md.txt#limitations).

## Private Service Connect endpoint deployment options

A Private Service Connect service attachment is generated from the
producer service (such as Gemini Enterprise Agent Platform). As a consumer, you can gain
access to the service producer by deploying a consumer endpoint in one or more
VPC networks.

## Deployment considerations

The following sections discuss considerations for communication from your
on-premises, multicloud, and VPC workloads to Google-managed
Gemini Enterprise Agent Platform services.

### Private Service Connect backends

Google does not support using
[Private Service Connect backends](https://docs.cloud.google.com/vpc/docs/access-apis-managed-services-private-service-connect-backends.md.txt)
with Gemini Enterprise Agent Platform online prediction endpoints.

### IP advertisement

- When you use Private Service Connect to connect to services in
  another VPC network, you choose an IP address from a [regular
  subnet](https://docs.cloud.google.com/vpc/docs/subnets#purpose) in your VPC network.

- By default, the Cloud Router will advertise regular VPC
  subnets unless custom advertisement mode is configured. For more
  information, see [Custom advertisement
  mode](https://docs.cloud.google.com/network-connectivity/docs/router/concepts/advertised-routes#am-custom).

- The IP address for the consumer endpoint must be in the same region as the
  service producer's service attachment. For more information, see [Service
  attachments](https://docs.cloud.google.com/vpc/docs/private-service-connect#service-attachments) and
  [Access published services through
  endpoints](https://docs.cloud.google.com/vpc/docs/configure-private-service-connect-services).

### Firewall rules

You must update the firewall rules for the VPC
network that connects your on-premises and multicloud environments
to Google Cloud to allow egress traffic to the
Private Service Connect endpoint subnet. For more information, see
[Firewall rules](https://docs.cloud.google.com/vpc/docs/manage-security-private-service-connect-consumers.md.txt#firewall-rules).

### ServiceAttachment reuse

> [!NOTE]
> **Note:** This section describes network resource reuse and is different from [model co-hosting](https://docs.cloud.google.com/gemini-enterprise-agent-platform/machine-learning/predictions/model-co-hosting.md.txt), which describes compute resource reuse with `DeploymentResourcePool`.

When multiple Gemini Enterprise Agent Platform prediction endpoints are configured to use
Private Service Connect with identical project allowlists, they might
share the same `ServiceAttachment` resource. This allows consumer projects in
the allowlist to connect to these prediction endpoints through the same
`ServiceAttachment`. This reuse isn't guaranteed.

If prediction endpoints use different project allowlists, they use different
`ServiceAttachment` resources to ensure that connections are only accepted
from projects on their respective allowlists.

When multiple prediction endpoints share a `ServiceAttachment`, we recommend
that you create and reuse one Private Service Connect endpoint for
that `ServiceAttachment`, rather than creating a separate
Private Service Connect endpoint for each prediction endpoint.