What is qwenproxy?

QwenProxy is a local server that lets you use the Qwen family of AI chat models through the same standard interface that OpenAI tools and SDKs expect. The core trick is that it does not call an official API. Instead, it automates a real web browser behind the scenes to interact with the Qwen chat website on your behalf, then translates those interactions into responses formatted like OpenAI's chat API. This means any app or script already written to talk to OpenAI's models can be redirected to QwenProxy with minimal changes, and it will transparently route requests through the Qwen models instead. The server runs locally on your machine and handles things like keeping your Qwen session alive so you do not have to log in repeatedly, supporting the reasoning or thinking mode some Qwen models offer, and running local tools as part of chat conversations. You configure it with your Qwen account credentials or log in once via browser, and then point your AI client software at the local address, for example http://localhost:3000/v1, instead of the usual OpenAI endpoint. A developer would use this when they want to experiment with Qwen models using existing OpenAI compatible code without paying for API access, or when they want to test different AI models by simply swapping the endpoint URL. The project comes with Docker support for easy deployment and is licensed under ISC. It is written in TypeScript, and the README states plainly that it is intended for educational and research purposes only, and does not endorse violating the Qwen platform's terms of service.