Deploy to Vertex AI Agent Engine¶
Google Cloud Vertex AI Agent Engine is a set of modular services that help developers scale and govern agents in production. The Agent Engine runtime enables you to deploy agents in production with end-to-end managed infrastructure so you can focus on creating intelligent and impactful agents. When you deploy an ADK agent to Agent Engine, your code runs in the Agent Engine runtime environment, which is part of the larger set of agent services provided by the Agent Engine product.
This guide includes the following deployment paths, which serve different purposes:
-
Standard deployment: Follow this standard deployment path if you have an existing Google Cloud project and if you want to carefully manage deploying an ADK agent to the Agent Engine runtime. This deployment path uses Cloud Console, ADK command line interface, and provides step-by-step instructions. This path is recommended for users who are already familiar with configuring Google Cloud projects, and users preparing for production deployments.
-
Agent Starter Pack deployment: Follow this accelerated deployment path if you do not have an existing Google Cloud project and are creating a project specifically for development and testing. The Agent Starter Pack (ASP) helps you deploy ADK projects quickly and it configures Google Cloud services that are not strictly necessary for running an ADK agent with the Agent Engine runtime.
Agent Engine service on Google Cloud
Agent Engine is a paid service and you may incur costs if you go above the no-cost access tier. More information can be found on the Agent Engine pricing page.
Deployment payload¶
When you deploy your ADK agent project to Agent Engine, the following content is uploaded to the service:
- Your ADK agent code
- Any dependencies declared in your ADK agent code
The deployment does not include the ADK API server or the ADK web user interface libraries. The Agent Engine service provides the libraries for ADK API server functionality.