A text box, streaming replies, and your own OpenAI key — running on an instance you own.
Yours in a click · runs behind your company login · free to try
One text box, one AI, nothing else to learn. Type a question, get an answer streamed back token by token, keep the thread going.
It runs on the OpenAI API with your own key, on your own instance — so the conversations stay in your database, not someone else's product.
What's inside
A single chat screen with streaming replies.
Conversation history that persists, so you can pick a thread back up.
Your OpenAI key in your own env vars. Swap the model whenever you want.
It's deliberately bare. That's the point — it's the smallest thing that works, and the easiest thing to build on. Ask for a system prompt that makes it a support agent, a sidebar of saved chats, file uploads, a different model per conversation. Start from simple, add only what you need.
This is the live demo, embedded right here. Click around; your copy starts out exactly like it.
You get your own copy — your data, your team, your changes. Ask for a tweak and it ships; switch it off any time.