FAQ
Frequently asked questions
Clear answers about sources, accuracy, privacy, access, and the current release.
Does Novus Learn require an account?
No. Public topics, search, preferences, and the local library work without login. This build has no authentication provider.
Does it use a paid AI API?
No. The current experience uses deterministic application code and public Wikimedia read APIs. No model key or secret is required.
Where does topic information come from?
Search and article text come from the selected Wikipedia edition. Structured identity comes from Wikidata when the article links an entity. Reusable media and licence metadata come from Wikimedia Commons.
Why do some topics say limited or ambiguous?
A disambiguation page represents several possible meanings and requires your choice. A limited-source state means the retrieved introduction or section structure is too sparse for a confident baseline. Novus Learn does not fill those gaps with guesses.
Where are saved topics stored?
Shortcuts use local storage, while generated source and knowledge snapshots use IndexedDB. Neither creates a cloud account. Clearing local data removes both.
Are analytics or ads running?
Google Analytics is available but off by default. It loads only if you accept the Analytics category in the cookie banner, and it uses IP anonymization. Advertising through Google AdSense is being prepared and stays off until you accept the Marketing category and an approved ad account is configured. Rejecting nonessential storage keeps both disabled.