seehiong/multi-model-chat — reverse-engineered prompt

Reverse engineered prompt

GitHub

Build me a clean web app where I can type one chat message and send it to several AI models at the same time, then compare all the answers side by side as they come in.

I want to be able to pick up to 5 models, including cloud models through OpenRouter and local models running with Ollama. Add a simple settings page where I can paste my API key, save it in the browser, and add or edit local model details. The chat should keep the layout easy to read, render markdown nicely, and show code blocks with syntax highlighting.

Please make it feel modern and responsive on desktop and phone. Use React, TypeScript, Vite, and Tailwind CSS. It should be easy to run locally with npm and easy to deploy on Vercel without needing server environment variables. Look up current docs online if you need to.

Want more depth? Deep Reverse