What “running AI locally” actually means.

The difference between an app on your computer and a model running on your computer.

Map the whole workflow

Model inference is only one component. Document conversion, search, embeddings, synchronization, and connected tools may operate elsewhere. If your objective is an offline workflow, identify each component and check its network behaviour. Do not assume that choosing a local model makes the entire application offline.

Continue with the guide

This topic is taught in one maintained lesson, with an example and a check you can try.

Run a local model and verify the actual setup