QwenLM/Qwen3 ? reverse-engineered prompt

I want to set up a local version of the new Qwen3 language model. Please create a project that lets me run one of the smaller but still powerful models on my own machine. The main thing I need is a simple web chat interface where I can try it out.

From what I've read, it has a couple of different modes. There's a special 'thinking mode' that's supposed to be amazing for complex stuff like reasoning, math problems, and writing code. Then there's a more general chat mode for creative writing and regular conversations. Build it so I can easily switch between these two modes to see the difference. It also mentions being able to process huge documents, so I'd love to be able to test that long-context feature as well. You can find the models on Hugging Face.