About this site

What this is

A free reference desk for the Bible, built around one main service: the Commentary. Name a book and a chapter and you get a narrative reading of it — where that chapter sits in the Bible's single storyline, what it gathers up from earlier chapters, and what later chapters do with it — one chapter at a time, so every chapter gets a full reading. Alongside the commentary, the directory answers direct questions — from "why does God allow suffering?" to a verse that's bothering you — with a plain-language answer, the exact passages to check it against, and, where Christians genuinely disagree, an honest map of the main positions instead of a pretended consensus. There are no accounts, no payments, and no tracking. The teaching pages — the Storyline and the Presence of God — walk the Bible's big picture the same way.

How the answers are made

Honestly, since you're right to ask. The answers come three ways, and each is labeled:

APPROVED LISTING — these were written with care, reviewed by a person, and approved before publication. They live in the Listings and don't change until they're re-reviewed.

GENERATED BRIEFING — if your question isn't in the Listings yet, the answer is generated live. This is not a general-purpose chatbot offering the internet's average opinion. The engine has been deliberately programmed with a specific narrative framework — a recovery of the Bible's own storyline, in which every question is answered from inside the one story Scripture tells, from the garden to the city. Its standing instructions bind it: answer within that story, cite the passages for every claim, present other traditions fairly, never pronounce on any person's standing with God, and say "this is genuinely debated" when it is. A human audit runs over what the line says; corrections tighten the instructions, and the strongest answers graduate into the Listings.

ANSWERED PREVIOUSLY — when a question has already been asked, the same generated answer is served again rather than re-made, so everyone asking gets an identical answer while it awaits the audit.

GENERATED COMMENTARY — the chapter readings on the front page are made the same way, under the same standing instructions and the same human audit. Each chapter is generated once, then kept — every reader of Genesis 3 sees the same reading — and a reading the audit approves is fixed permanently, while one it rejects is removed and remade under tightened instructions.

So: does the AI "only pull from Scripture"? The straight answer: the engine is a language model — it has read far more than the Bible, and no honest site would claim otherwise. What makes this site different is that the model is not left to its own judgment. It is programmed to read and answer inside the Bible's own recovered storyline, and it is required to show its work — every answer carries the passages it rests on, so you can check it for yourself.

One more honesty note: where an answer leans on a historical claim — who taught what, and in which century — Bible passages can't verify that kind of statement. Those claims rest on the historical record, and as this site matures they are being footnoted to primary sources and mainstream scholarship. If you catch one that needs its footnote sooner, you've found exactly the kind of correction this site wants.

Why this isn't a search-engine answer

Ask a search engine's AI about the Bible and you get an average — the internet's most-repeated opinions, blended fresh each time, slightly different with every asking, and owned by no one. This site is built on the opposite premise.

The answers here come from one deliberate reading, not a blend. Behind the engine sits a framework built over years of work, which reads the Bible as a single story told in its own vocabulary — covenant, presence, promise, exile, return — rather than through whatever the internet happens to say most often. And that reading is not a novelty: it recovers how the earliest readers heard these texts, and where it takes a position, it can show ancient company for it. What a search engine averages is mostly the loudest voices of the last few centuries; recovering the older frame is precisely the point.

The difference shows up in practice. The rules here are fixed and purpose-built — no personalization, no engagement optimization, nothing tuned to keep you scrolling. The same question earns the same answer, and the strongest answers are published openly in the Listings where anyone can read them. Every claim must carry its passages. Contested questions are called contested instead of being averaged into a false consensus. And a person audits this line and answers for it. An overview has no owner. This shelf does.

Scripture quotations

The chapter’s own text sits beside every reading in the Berean Standard Bible (BSB), quotations follow its wording, and every passage reference on the site links to the full BSB text so you can read it in context — checking is the point.

What this site is not

It is not a church, a pastor, or a counselor — and it will tell you so. Questions about faith belong finally in a room with people, and every answer here ends by pointing toward a congregation. If you're in crisis, this site steps out of the way entirely and points you to real help (in the US, call or text 988, any hour).

Your questions and your privacy

Questions you ask may be stored and reviewed to improve the service — that's how answers get audited and better. Please don't include names or private personal details in a question. There are no accounts, no advertising profiles, and no behavioral tracking; our hosting provider collects basic, anonymous performance metrics so the site stays fast.

For AI agents

This site keeps a second door open for software agents: a Model Context Protocol (MCP) server at mcp.thenearword.com/mcp. It serves the same shelf — approved listings, previously answered questions, and chapter commentary — plus an open orientation document on the Bible's storyline, and every payload is labeled for what it is, so an agent can tell its reader exactly which shelf it is quoting from. If you operate an agent, point it at that endpoint; a machine-readable summary of this whole site lives at /llms.txt.

Early days

This site is young and in open testing. The Listings are few and growing; the answers improve as readers push on them. If something seems off, the best thing you can do is what the Bereans did — check the texts — and the second-best is to ask the question again a better way. Both make the shelf truer.