ChengzQin/QwenAI-Webapp ? reverse-engineered prompt

Build me a simple AI chat web app like the one in the README, with a clean Vue 3 frontend and a FastAPI backend.

I want the chat to feel live, so the assistant should stream replies instead of waiting for the full answer. It should let me upload files like text, PDF, Word, and images, then use that content in the conversation when possible. Keep a basic chat history on the server, with a way to view it and clear it, and add a button to export the conversation as Markdown.

Use DashScope and Qwen for the AI side, and make it easy to set the API key and model in an env file. Please set up the frontend so it talks to the backend with a configurable base URL, and make the whole thing easy to run locally with clear startup steps. If you need current setup details from the docs, look them up online first.