Get help

The model gives a slow or strange first answer

The first message after loading a model is always the slowest — the model is warming up. If replies stay slow, open the Model library and pick a smaller model; the fit dot next to each one tells you what your machine runs comfortably. If an answer loops or comes back empty, the app says so and suggests the fix — usually rephrasing, or a larger model.

Which model should I download?

The one the app recommends — the recommendation is computed from your RAM, not from marketing. If you have disk to spare, download several: with “Suggest the best model” on, Quenderin proposes the right installed model per task (code questions go to the coder model, and so on). It suggests; it never switches silently.

Where are my chats? Can you see them?

On your device, and no. There is no account, no server, no telemetry. Deleting the app deletes everything. That's the entire architecture — the privacy policy is short because there's nothing to disclose.

How do I free up disk space?

Every installed model shows its on-disk size in the Model library and in Settings → Storage; right-click (or use the ⋯ menu) to delete it or reveal the file. Models re-download any time.

Understand what you're running

Quenderin runs real AI models on your own machine — no cloud, no account, nothing leaves the device. It's worth knowing the honest trade-offs and where this is all headed:

Something else, or something broken?

← Back to Quenderin · FAQ · Changelog · Privacy