Use Agent Development Kit to Build, Test, and Deploy Agentspublic

The Agent Development Kit (ADK) is an SDK to build, test, and deploy agent workflows from within your development environments. You can opt in the public preview from the Feature Preview page.

You must have the following prerequisites to use the ADK:

• Python version 3.13

• Dependencies listed in requirements.txt at the root of the folder or repo to deploy.

• .env file with environment variables to use in agent deployment.

• Model access key for authentication. Set the key in the GRADIENT_MODEL_ACCESS_KEY environment variable and add it to your .env file. For running your agent locally, you must export the key in your terminal for it to be accessible to the application.

• Your account’s personal access token. The key must have all CRUD scopes for genai and read scope for project. Set the API key in the DIGITALOCEAN_API_TOKEN environment variable and add it to your .env file to enable deploying the agent to your DigitalOcean account.

To build and deploy a new agent using the ADK, follow these steps:

1. Install the ADK using the following command:

   pip install gradient-adk

   Installing the ‘gradient-adk’ package automatically gives you access to the ‘gradient’ CLI.

2. Initialize a new project using the following command:

   gradient agent init

   To provide an easy way for you to get started, the command creates folders and files, and sets up a base template (main.py) for a simple LangGraph agent that makes calls to a openai-gpt-oss-120b model using serverless inference. When prompted, specify an agent workspace name and an agent deployment name.

3. Run and test the example agent locally using the following command:

   gradient agent run

   You can then access the agent and interact with it at the http://0.0.0.0:8080/run endpoint.

   To interact with the agent, send a POST request to this endpoint with a prompt in the request body. For example, your request body can be '{"prompt": "How are you?"}'. Your agent processes the request and returns a response.

4. Deploy your agent using the following command:

   gradient agent deploy

   After the deployment succeeds, you can see the deployment URL that the agent is running on in your terminal.

   Test the deployed agent by sending a POST request with a prompt in the request body to the URL. For example, the request body can be '{"prompt": "hello"}. Your agent deployment processes the request and returns a response.

For more detailed information on building, testing, and deploying agents using the ADK, see Build Agents Using Agent Development Kit.