Handle edge cases in import

**Feature request**
It would be great to have caching for models.

**Use case**
Many users need this functionality for production deployments.

**Implementation idea**
Add a new config option `enable_caching` with default `false`.

No files, tests, or entry points are named. Start by tracing how model imports and configuration are handled, then clarify the caching behavior and verify the default-disabled option in the relevant tests.

Written by the indexing model from the issue text.