About
This gateway started as a weekend project: I wanted one stable endpoint for my own tools to talk to Google's Gemini models, without wiring API keys into every script and app. It relays requests, translates them into the format Gemini expects, and sends the answers back.
It runs on a small VPS, uses an OpenAI-compatible interface so existing clients work unchanged, and is used only by me and a few personal projects. It is not a business, there are no accounts, and I don't collect or store prompts.
Contact
There is no support channel — it's a one-person project. If something breaks, it gets fixed when I notice it.
Privacy
Requests are relayed, not logged. No prompts, responses, or API keys are stored on this server beyond what the gateway needs in memory to complete a request.