Agent Garden - Samples for learning, discovering and building

Today, we’re expanding access to Agent Garden by making it available to all users, not just those who use Google Cloud. Creating AI agents, particularly sophisticated multi-agent systems, may be difficult. Agent Garden helps address this by mitigating a primary obstacle developers encounter in the design of the agent itself. This process demands extensive investigation, precise coordination, and substantial developmental investment to integrate diverse tools and frameworks.

Agent Garden can help developers with challenges such as:

• Developing AI agents that address complex business challenges
• Creating intricate multi-agent workflows that blend generative AI with deterministic logic
• Deploying and integrating agents seamlessly with existing systems and data sources

Agent Garden offers a growing repository of curated agent samples, solutions, and tools, designed to accelerate the development and deployment of powerful AI agents using the Agent Development Kit (ADK), Agent Garden provides a unique learning experience and resources to jump start your agent development journey.

Discovering Agent Garden's Capabilities

Agent Garden offers a robust collection of features tailored to various use cases. These code samples, developed with ADK, provide extensive integration with cloud services like BigQuery and Vertex AI Search, among others. We highly recommend exploring the available resources to understand their full potential.

Every agent sample includes a detailed overview, applicable use cases, architectural insights, and a description of its capabilities. Developers can also access the sample code directly on GitHub for in-depth review.

Deploy sample agents with one click

In addition to learning the code, Agent Garden now helps developers deploy agents with Agent Starter Pack (open source starter pack that helps with seamless deployments to production). Developers can now simply click deploy and sample agents are deployed to Agent Engine in your project and exposed through the Agent Engine playground UI. You can test and experiment with almost no effort.

Customizing the agent and make it your own

Most developers intend to customize their code to fit their needs and use cases. We provide an option through Firebase Studio so you can open and customize the agents

The above developer journey to explore, learn, deploy and customize is intended to make ADK agent development seamless for your use case.

How early testers have used it

Customers are increasingly leveraging the innovative samples provided, demonstrating successful adoption and tangible business value. A prime example of this success is Renault Group, which recently shared its compelling journey during Google Cloud Next’2025. They detailed how they integrated a sophisticated data scientist agent into their electric vehicle (EV) charger platform. This integration has significantly enhanced their operations and user experience by giving the business team autonomy to directly leverage their data.

Frequently Asked Questions:

• Can these agents be deployed in production directly?
  • Both adk-samples and agent starter pack are open source software subject to the terms set forth in the applicable source code.
• Can the deployed agents be accessed and shared?
  • Agents are deployed in Agent Engine and are subject to the access management options of Agent Engine
• How can I delete the deployed agent?
  • Agents are deployed in Agent Engine and can be deleted within Agent Engine
• Where do I look for more samples?
  • Agent Garden will feature curated examples of high-impact use cases, with additional samples available at ADK samples, ADK community repo and the ADK docs

To stay actively engaged with the community, please do participate in the ADK community

Start Building Today

Resources:

• Agent Garden
• Agent Starter Pack
• Agent Development Kit -> ADK Docs, ADK samples, ADK Python community