Two of these generate prose and one returns the passages the prose would have been built from. For research — as opposed to devotional reading — the raw retrieval is the honest one: you see what the corpus actually says and judge it yourself. The theological position is declared up front rather than implied, which is the right way to publish something like this. No account, no stored data, and a rate limit as the only gate.
BEREAN.AI's endpoint, open and rate-limited by IP. We connected unauthenticated and it listed five tools — two answer questions, one searches the corpus directly, two return the day's readings.
- ask_question — a concise pastoral answer of 250-400 words grounded in Reformed theology
- scholar_query — academic research using two-stage retrieval: dense vector search followed by cross-encoder reranking
- search_sources — the knowledge base itself, returning passages without a generated answer
- get_daily_devotion — today's Scripture reference, verse text and devotional message
- get_daily_exposition — the One Year Bible reading with its exposition
None. No account, and the vendor states no personal data is collected — there is nothing stored to opt out of.
Paste a URL — add the endpoint to your client