FAQ
Short answers to the questions that come up before people start.
Is this just for scholars?
No. It is built for anyone who takes the text seriously — a first-time reader with a question their study Bible did not answer, and a professor preparing a lecture use the same search box. The output is research-grade; the interface is a search bar.
Does it take a side?
Where the traditions disagree, the analysis says so and names who holds which reading. It does not adjudicate. For passages from the Hebrew scriptures it treats the text as the Tanakh and engages Jewish interpretation explicitly, rather than reading the Old Testament as background to the New.
There is a position baked in, though: that a passage should be read in its historical and political setting rather than as a free-floating proof text. Every analysis spends most of its length on that setting.
Which texts are included?
The Tanakh / Old Testament (23,145 verses), the New Testament (7,957), the Quran (6,236 ayat, with Arabic and tafsir), and 432 cross-referenced Hadith narrations from Sahih al-Bukhari, Sahih Muslim, and Sunan Abu Dawud, each with its authenticity grade. 6 English Bible translations. Hebrew for the Old Testament. See Translations and original languages.
Do I need an account?
Yes, and right now it has to be an invited address — TriCanon is in beta. Join the waitlist to get one. There is no password: the email address is the login.
Does it work offline?
No. Analysis and chat call a live research model on every request. Without a connection the app shows a “no connection” page and reloads itself when you are back. Your saved sessions are on the server and are not lost.
Is there an app?
It installs to your home screen from the browser — full screen, own icon, no app store. See Install the app.
How much does it cost?
Search, reading, translations, and original languages are free, with 10 AI analyses a month. Explorer and Scholar are announced but not on sale, and no prices have been published. See Plans and limits.
How reliable is the analysis?
Every analysis is grounded in at least three web sources that the model actually retrieved, and the sources are listed so you can check them. The prompt forbids inventing citations and requires dropping claims it cannot ground. That makes it dependable as a starting point and a bibliography — not as a final authority, and not as a substitute for the primary text or peer-reviewed work.
The honest failure mode is subtle: a sound summary built from a weak source. Reading the citation list catches most of it.
Can I save or export my research?
Every analysis is saved to your account automatically. Export (PDF, Markdown) is planned for the Scholar tier and not shipped yet.
Does it read Hebrew, Greek, and Arabic?
It shows Hebrew for Old Testament verses (with the JPS translation selected), Arabic for the Quran with tafsir, and Arabic plus English for Hadith. Hebrew and Greek word definitions with their semantic range are a Scholar-tier feature, not yet available.