Bring Your Own AI Backend | Loopal
Frequently asked questions
- Can I use my own API key instead of platform credits?
- Yes. Bring your own Anthropic, OpenAI, Doubao or Gemini key, or point it at your own Ollama server.
- What happens if I do nothing?
- It uses platform credits out of the box, so it works without any setup.
- What does running my own backend change?
- Your requests go where you chose rather than to a provider the app selected for you.
- What can it do with permissions?
- Read your schedule, create reminders, and answer location-aware questions — each capability can be turned off on its own.
The short answer
Use platform credits out of the box, bring your own Anthropic / OpenAI / Doubao / Gemini key, or point it at your own Ollama server.
That third option is the one that matters most, and it is rare in this category.
Why "your own backend" is the real feature
Platform credits are the easy path, and they mean the model provider was chosen by the app rather than by you. Pointing Loopal at an Ollama server on your LAN means the request never leaves your network at all.
For a keyboard — the app you type every private thought into — that is a structural property, not a preference.
Zero setup if you want it
If you do nothing, it works on platform credits. The choice of backend is available, not required.
Permissions, individually
With your permission it can read your schedule, create reminders and answer location-aware questions — and every capability can be turned off on its own.
Worth noting the granularity: these are separate switches, so granting schedule access does not imply granting location.