Frequently asked questions - gpu-snap
How it works
- Sign in and top up your balance.
- Press one button - a preconfigured GPU virtual server starts for you.
- Use the private AI chat in your browser or call the LLM API endpoint from your code. Stop when you are done.
Private by design
- A dedicated GPU server for you only, not a shared model service.
- No prompt or chat history is saved.
- After stop the server is deleted together with all its data.
Pay for the time, not for the hassle
On-demand GPU compute billed only for the time the server runs. No subscription, no setup, unused time is returned when you stop early.
Frequently asked questions
Are my prompts or chats stored?
No. Prompts and chat history are not saved, and the server is yours alone while it runs.
What happens when I stop the server?
The GPU server is stopped and deleted together with everything on it.
Can I use it as an LLM API endpoint?
Yes. Every server exposes a private LLM API endpoint in addition to the browser AI chat.
How do I pay?
Top up your balance once and pay only for the minutes your GPU server runs.
How fast does the server start?
A preconfigured GPU server is usually ready within a few minutes after one click.