The source does not leave for a model
Deterministic generation runs without any AI provider in the path. A model is contacted only for a generation where you selected Luna and confirmed the disclosure for that source.
Study material you can trace back to the sentence it came from.
The private engine segments your source and builds cards from sentences inside it. No model provider is contacted, no AI allowance is consumed, and the same verbatim citation contract applies. It is the default path, not a stripped-down mode you have to hunt for.
Paste any notes below and the page itself drafts a few anchored cards with the same private engine the product saves with — no account, no upload, nothing leaves this browser tab.
This preview runs entirely in this browser tab, powered by the same private deterministic engine the product uses — no model provider is involved and nothing you paste here is uploaded or stored.
What 'private' means here, concretely.
Deterministic generation runs without any AI provider in the path. A model is contacted only for a generation where you selected Luna and confirmed the disclosure for that source.
Decks, schedules, review history, quiz attempts, and generation records are stored against your own Google-authenticated account. There is no public deck directory and no shared library.
The same source produces the same segmentation. Each deck stores a SHA-256 hash of its source, so you can tell whether two decks came from identical material.
Export a deck or the entire account as JSON, or delete everything — cloud library, generation records, profile, and the browser recovery copy — from Settings.
The private path, start to finish.
Google sign-in creates the private workspace. Paste up to 30,000 characters, or upload a .txt or .md file up to 1 MB.
Leave the private engine selected and generate. No consent box appears because nothing is being sent to a provider.
Review on schedule, export JSON at any time, and delete individual decks or the whole account whenever you want.
Being precise about privacy:
Not to any AI model provider. The source is processed by RecallQuill and stored in your own account so your library works across devices. A model is involved only in a generation where you chose Luna and confirmed the disclosure.
Yes. The same contract applies: every card records the paragraph it came from and a verbatim quote from that paragraph, and the same relationship checks run before anything is saved.
It phrases prompts more mechanically, since it works from sentence structure rather than meaning. In exchange it is instant, unlimited, repeatable, and never consumes your daily AI allowance.
Settings has a delete action that removes the cloud library, review history, generation records, profile settings, and the browser recovery copy, then deletes the authentication record itself.