response = client.chat.completions.create(
                model="gpt-4o-mini",
                messages=messages,
                max_tokens=300
)