A study tool, not another AI app.

We want more learners to have access to effective study tools. Our long-term goal is to lower subscription fees as our costs fall sustainably, while preserving educational quality and a reliable service.

With sufficient financial and operational resources, we hope to run suitable AI models on infrastructure we control, reducing our dependence on third-party AI APIs. We would migrate only if evaluations on our actual study tasks show quality that meets or exceeds our current providers, and reliability and total operating costs make sense.

Self-hosting means running AI on our infrastructure. It does not mean on-device or offline AI, or training a proprietary model. Today, our AI features still use external providers, as described in our Privacy Policy.

These are long-term goals, with no promised timeline or price reduction. A subscription provides the benefits shown at checkout; future capabilities are not guaranteed included benefits.

AnkiCat is not an Anki add-on. It is not a cat theme pack, and it is not affiliated with the Anki project. It is a standalone study app. Both use spaced repetition as a technique, but AnkiCat is its own product with its own apps.

The whole loop is ours: drafting cards from your material, scheduling reviews, quizzes, analytics, and the native iOS and Android apps. One team owns the whole experience end to end, so nothing falls between vendors. When something breaks, exactly one team fixes it.

AI is just one small part of AnkiCat

AI drafts study content from your material. Quiz answers are scored against stored answers. A Student Pro subscription includes a limited number of AI credits. Extra packs are not for sale. Subscribe in the app. The store shows the price in your currency.

We keep generative AI optional and visible on purpose. Data centers that run it use real compute, electricity, and water. So it does not run in the background across the app.

You choose when to use it, and you always know where it runs.

AnkiCat is dark-first because people study at night. It works offline because memory does not wait for a connection. It collects the minimum data needed to run the service: no feed, no fake urgency, no engagement tricks.

The app wins by being good at one job: helping you remember what you learned. If it does that job well, you will come back on your own.