import os
import dashscope
dashscope.base_http_api_url = "https://dashscope.aliyuncs.com/api/v1"

messages = [
    {
        "role": "user",
        "content": [
                    {"text": "图中描绘的是什么景象?"}]
    }]
response = dashscope.MultiModalConversation.call(
    api_key=os.getenv('DASHSCOPE_API_KEY'),
    model='qwen3.7-plus',
    messages=messages
)
print(response.output.choices[0].message.content[0]["text"])