from agents import AsyncOpenAI, OpenAIChatCompletionsModel
from agents.run import RunConfig
external_client = AsyncOpenAI(
    api_key='1234',
    base_url="http://localhost:11434/v1",
)
model = OpenAIChatCompletionsModel(
    model="qwen3:0.6b", openai_client=external_client
)
config = RunConfig(model=model, tracing_disabled=True)
# This configuration is for ollama!