Open In Colab Open on GitHub In the Wikipedia Search Tools notebook, we demonstrated how to create Agents with basic Wikipedia search capabilities. Now, we’re taking it a step further with WikipediaAgent—a powerful agent equipped with built-in wiki search tools right out of the box!

​
Package Installation

To get started with the Wikipedia search integration in AG2, follow these steps: Install AG2 with "wikipedia and openai since we use OpenAI’s LLMs in our example: 
pip install -U "ag2[wikipedia, openai]"
Note: If you have been using autogen or ag2, all you need to do is upgrade it using: 
pip install -U "autogen[wikipedia, openai]"
or 
pip install -U "ag2[wikipedia, openai]"
as autogen, and ag2 are aliases for the same PyPI package.
You’re all set! Now you can start using Wikipedia Search with AG2.

​
Demonstration

The WikipediaAgent comes packaged with two Wikipedia tools, WikipediaQueryRunTool for running title/keyword searches and WikipediaPageLoadTool for fetching full page content. Together with the fact that it has a pre-configured system message (which you can override), you only need to create the agent and add it to a chat to get going.

​
Imports

from autogen import LLMConfig
from autogen.agents.experimental import WikipediaAgent

​
Agent Configuration

Configure WikipediaAgent and run it. Using the agent’s run command will automatically 
# LLM configuration
config_list = LLMConfig(api_type="openai", model="gpt-4o-mini")

# Create the agent
wiki_agent = WikipediaAgent(name="wiki-agent", llm_config=config_list)

# If you are adding this agent into a chat that requires separate tool execution, register the tools with the executing agent:
# for tool in wiki_agent.tools:
# tool.register_for_execution(the_executing_agent)

​
Start the Conversation

With the setup complete, you can now use the conversation. 
response = wiki_agent.run(
    message="What's the population of Australia?",
    max_turns=2,
    # Instruct the chat to associate the WikipediaAgent's tools with the internal tool executor
    tools=wiki_agent.tools,
)

# Run the chat
response.process()
Edit this page