Gemini ADK Tutorial

A collection of example agents built with Google's Agent Development Kit (ADK).

Prerequisites

• Python 3.13+
• Google API Key (for Gemini models)
• Optional: OpenAI API Key (for OpenAI model examples)
• Optional: Tavily API Key (for travel agent)

Installation

# Clone the repository
git clone https://github.com/alfredang/gemini-adk-tutorial.git
cd gemini-adk-tutorial

# Install dependencies using uv
uv sync

# Or using pip
pip install -e .

Configuration

Create a .env file in each agent directory with your API keys:

GOOGLE_GENAI_USE_VERTEXAI=0
GOOGLE_API_KEY=your-google-api-key
OPENAI_API_KEY=your-openai-api-key  # Optional, for OpenAI examples
TAVILY_API_KEY=your-tavily-api-key  # Optional, for travel_agent

Running Agents

Use the ADK CLI to run any agent:

adk run <agent_directory>

For example:

adk run my_agent
adk run stock_agent
adk run travel_agent

Agent Examples

my_agent

A basic "Hello World" agent using Gemini.

from google.adk.agents import Agent

root_agent = Agent(
    model='gemini-2.5-flash',
    name='root_agent',
    description='A helpful assistant for user questions.',
    instruction='Answer user questions to the best of your knowledge',
)

my_agent_model

Demonstrates using OpenAI models via LiteLlm integration.

from google.adk.agents import Agent
from google.adk.models.lite_llm import LiteLlm

root_agent = Agent(
    model=LiteLlm(model="openai/gpt-4.1-mini"),
    ...
)

my_agent_tools

Shows how to add tools (like Google Search) to an agent.

from google.adk.agents import Agent
from google.adk.tools import google_search

root_agent = Agent(
    model='gemini-2.5-flash',
    tools=[google_search],
    ...
)

my_agent_session

Demonstrates session management with InMemorySessionService and Runner.

stock_agent

A hierarchical multi-agent system for stock analysis:

• root_agent - Greets user and introduces capabilities
• stock_workflow_agent (SequentialAgent) - Orchestrates the workflow
  • ticker_input_agent - Collects stock ticker from user
  • stock_research_agent - Researches and generates a detailed report

travel_agent

A multi-agent travel planner with specialized sub-agents:

• planner_agent - Creates day-by-day itineraries
• budget_agent - Estimates travel costs
• local_guide_agent - Provides food and cultural tips
• research_agent - Searches for current travel information (uses Tavily)

tutor_agent

A multi-agent tutoring system with subject-specific tutors:

• math_tutor_agent - Helps with mathematics
• physics_tutor_agent - Helps with physics
• history_tutor_agent - Helps with history

Key Concepts

Agent Types

Agent Type	Description
Agent / LlmAgent	Basic LLM-powered agent
SequentialAgent	Runs sub-agents in sequence
ParallelAgent	Runs sub-agents in parallel
LoopAgent	Runs sub-agents in a loop

Using Different Models

# Gemini (default)
model='gemini-2.5-flash'

# OpenAI via LiteLlm
from google.adk.models.lite_llm import LiteLlm
model=LiteLlm(model="openai/gpt-4o-mini")

# Claude via LiteLlm
model=LiteLlm(model="anthropic/claude-3-5-sonnet-20241022")

Adding Tools

from google.adk.tools import google_search

# Built-in tools
tools=[google_search]

# Custom function tools
def my_tool(query: str) -> dict:
    return {"result": "..."}

tools=[my_tool]

Sub-agents

root_agent = Agent(
    name="root_agent",
    sub_agents=[agent1, agent2, agent3],
    ...
)

Resources

• Google ADK Documentation
• Google ADK GitHub
• LiteLlm Documentation

License

MIT