Skip to content
Open
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
26 changes: 24 additions & 2 deletions docs/deploy/agent-engine/deploy.md
Original file line number Diff line number Diff line change
@@ -1,7 +1,7 @@
# Deploy to Vertex AI Agent Engine

<div class="language-support-tag" title="Vertex AI Agent Engine currently supports only Python.">
<span class="lst-supported">Supported in ADK</span><span class="lst-python">Python</span>
<div class=\"language-support-tag\" title=\"Vertex AI Agent Engine currently supports only Python.\">
<span class=\"lst-supported\">Supported in ADK</span><span class=\"lst-python\">Python</span>
</div>

This deployment procedure describes how to perform a standard deployment of
Expand Down Expand Up @@ -146,6 +146,28 @@ For `region`, you can find a list of the supported regions on the
To learn about the CLI options for the `adk deploy agent_engine` command, see the
[ADK CLI Reference](https://google.github.io/adk-docs/api-reference/cli/cli.html#adk-deploy-agent-engine).

### Deploy with an API key (Express Mode)

If you don't have a Google Cloud project set up, you can still deploy your
agent to Agent Engine using **Express Mode**, which authenticates using an
API key. This is a quicker way to get started if you don't need the full
capabilities of a Google Cloud project.

To deploy in Express Mode, use the `--api_key` option with the
`adk deploy agent_engine` command:

```shell
API_KEY=my-api-key

adk deploy agent_engine \
--api_key=$API_KEY \
--display_name="My First Agent" \
multi_tool_agent
```

When you use the `--api_key` option, the ADK initializes the Vertex AI SDK
with your API key instead of using project and location credentials.

### Deploy command output

Once successfully deployed, you should see the following output:
Expand Down
Loading