Skip to content
Nuci

Draft. A lawyer has not reviewed this page yet. It is not legal advice.

About the artificial intelligence

Nuci uses an artificial intelligence system. You are not talking to a person.

What the AI does

It proposes questions from the text you write, checks whether your answers contain what matters, plays a curious classmate in depth tasks and annotates your living notes.

Questions come only from your text. If the AI finds something missing, it asks you instead of making it up, and each question shows the sentence of your text it comes from.

What it does not do

It does not decide when each question comes back: a memory model computes that from what you recalled.

It does not write your notes for you or rewrite yours.

Which models

Nuci shows its AI models by short names. Behind each name is a Claude model made by Anthropic:

Passer is Claude Haiku 5.5. It grades every memory test, on every plan.

Falco is Claude Sonnet 5.5. On every plan it proposes questions from your notes and reviews a grade when you ask. On Pro and during the free trial it also runs depth tasks, living notes and diagrams.

Aquila is Claude Opus 5.5. It runs depth tasks, living notes and diagrams on Ultra, and on topics marked complex in Pro. It does not propose questions.

Diagrams

In mathematics and science, the AI may describe a diagram as data: points, lines, angles and labels. Nuci checks that description and draws it. The AI does not draw, and a diagram can still be wrong.

It can be wrong

If you think a grade is unfair, select “Review my grade” and a more capable model checks it again. Always read the answer Nuci shows you.

Your data and the AI

We send the model provider only what each task needs. We do not use your answers to train models. More detail on the privacy page.