def compare(images: List[str], prompt: Optional[str], detail: Optional[str], model: Optional[str]) -> str

agent = VisionAgent()
    comparison = agent.compare([
        "before.jpg",
        "after.jpg"
    ], prompt="What changed between these images?")