Get up and running with AG2 in just 3 minutes! This guide will help you set up your environment and build your very first multi-agent workflow. In just a few steps, you’ll have your first agent up and running. Let’s make it happen!

​
Set Up Your Environment

We recommend using a virtual environment for your project to keep your packages contained. See venv.

Install AG2

AG2 requires Python version >= 3.9, < 3.14. Install AG2 with OpenAI integration using pip:

pip install ag2[openai]

The package is available under ag2, pyautogen, or autogen names. The default installation includes minimal dependencies, you can add extra options based on your specific requirements.

From version 0.8: The OpenAI package, openai, is not installed by default.

Install AG2 with your preferred model provider(s), for example:

  • pip install ag2[openai]
  • pip install ag2[gemini]
  • pip install ag2[anthropic,cohere,mistral]

On Mac OS, if you get “no matches found:”, add a quote to the package name, for example:

  • pip install "ag2[openai]"

​
Build Your First Agent Workflow

Let’s build a poetic AI assistant that responds in rhymes using AG2 and OpenAI’s GPT-4o-mini model.

This example demonstrates how to:

  • Set up an LLM configuration
  • Create a conversational AI agent
  • Run an interactive multi-turn conversation

Create a Python script called first_agent.py, and paste the following code into it:

# 1. Import our agent class
from autogen import ConversableAgent, LLMConfig

# 2. Define our LLM configuration for OpenAI's GPT-4o mini
#    uses the OPENAI_API_KEY environment variable
llm_config = LLMConfig(api_type="openai", model="gpt-4o-mini")

# 3. Create our LLM agent
with llm_config:
    my_agent = ConversableAgent(
        name="helpful_agent",
        system_message="You are a poetic AI assistant, respond in rhyme.",
    )

# 4. Run the agent with a prompt
response = my_agent.run(
    message="In one sentence, what's the big deal about AI?",
    max_turns=3,
    user_input=True
)

# 5. Iterate through the chat automatically with console output
response.process()

# 6. Print the chat
print(response.messages)

???+ info

Why Two Steps: run() and process()

You might wonder why we need to call both run() and process() to get results.

👉 Here’s what’s happening:

When you call run(), it doesn’t immediately give you the final output. Instead, it returns an iterator, a special object that holds a stream of events, messages, and metadata.

👉 Why? Because flexibility matters.

The workflow steps won’t actually start running until you iterate over this iterator. This design gives you full control and makes it easy to build things like:

  • Custom UIs
  • Real-time dashboards
  • Interactive apps where you control how and when each event is handled

👉 Okay — so what does process() do?

process() is a built-in helper method that takes care of iterating through those events for you. It simulates a chat-like console experience — printing messages, handling user inputs, and making it feel like a live conversation.

👉 In short:

Use run() when you want full control over the workflow’s events and outputs.

Use process() along with run() when you just want a quick, ready-to-go chat experience in the console.

👉 Learn more about the run() method here →

​
Run Your Example

Now you’re ready to see your poetic AI agent in action!

Before running this code, make sure to set your OPENAI_API_KEY as an environment variable. This example uses gpt-4o-mini, but you can replace it with any other model supported by AG2.

=== “macOS / Linux”

export OPENAI_API_KEY="YOUR_API_KEY"

=== “Windows”

setx OPENAI_API_KEY "YOUR_API_KEY"

In your terminal, run:

python first_agent.py

If everything is set up correctly, the agent will reply to your initial message in rhyme, then prompt you for a response. You can either:

  • Type a reply — and the agent will respond in rhyme to your message
  • Press Enter — to send an empty message to the agent, and see how it creatively responds
  • Type exit — to end the conversation

The interaction continues for up to 3 turns (or until you exit).

​
Example Output

user (to helpful_agent):

In one sentence, what's the big deal about AI?

--------------------------------------------------------------------------------

>>>>>>>> USING AUTO REPLY...
helpful_agent (to user):

AI transforms our world, enhancing life’s parade,
With insights and solutions, it helps plans cascade.

--------------------------------------------------------------------------------
Replying as user. Provide feedback to helpful_agent. Press enter to skip and use auto-reply, or type 'exit' to end the conversation:

That’s it—you’ve built your first multi-agent system with AG2 🎉

Edit this page