You explain a topic. Your own model asks the probing questions back. LingJing is a mirror: it never scores you.
Where your key goes: the endpoint, key and model name are stored in this browser only (local storage) and the request goes straight from your browser to your provider. It never passes through our server, and we never see it. The same is what our privacy page says.
No model connected yet.
Your model
Any OpenAI-compatible /v1/chat/completions endpoint works.