Our vision
This page says why we built Glimsy and how we think about studying with flashcards, and below it answers the questions that come up most often along the way.
Studying with flashcards comes down to one thing: recalling the answer before you see it, and coming back to it right when it starts to fade. Everything else — screens, settings, syncing — is plumbing around that one mechanism. Glimsy exists because in the tools we used ourselves, the plumbing took more attention than the studying did.
The problem fitted into a single sentence: to study with a good review algorithm, you first had to work through a dozen screens of settings that nobody explains. We decided that a decision a user cannot make in an informed way should not be shown at all — rather than shown with a longer help text next to it.
The second thing we cared about is that the app should genuinely work out when you need to see a card again. A set of flashcards you page through end to end, marking each one “got it” or “not yet”, will show you the same cards in the same order next time — it has no way of knowing that this one card fades faster than the rest. In Glimsy the dates are computed by FSRS, an algorithm that estimates for each card separately how quickly it is fading, and that is what decides what you see today.
Where this is going. The things we are working on next all serve the same idea: take off the learner work that is not studying. Hence the Anki import that carries your progress over, hence an assistant that builds a deck from a topic or from material you upload.
Below we answer the questions we get most often — about flashcards themselves, about choosing an app, about your own material, and about studying before an exam.
Flashcards, and how to study with them
A flashcard is suspiciously simple: a question on one side, an answer on the other. All of its value comes from what happens in the second in between — the attempt to recall. That is why two people with the same deck can end up with completely different results: one is trying to remember, the other is flipping through.
What are flashcards, and why do they beat rereading notes?
A flashcard forces you to try to recall the answer before you see it. A page of notes does not — reading it, you recognise text you already know, and recognition is easy to mistake for being able to reproduce something. The difference is measurable: material tested with questions comes back more reliably than the same material read one more time, and the harder the recall was, the more durable the trace it leaves.
The practical conclusion is an uncomfortable one: studying with flashcards should feel harder than reading.
What is spaced repetition and why bother with it?
Memory does not fade evenly — most of the loss happens in the first hours and days after studying, and then the curve flattens out. Spaced repetition means returning to material just before the point where it starts to blur: that is when a single recall does the most work and pushes the next date furthest away. In practice a well-established card comes back once every few months, while the one you keep struggling with comes back tomorrow.
The date is computed by an algorithm. Today’s standard is FSRS, which instead of fixed multipliers estimates for each card separately how quickly you forget it. Glimsy uses exactly that and sets the dates itself — there is no field here where you type an interval.
How many flashcards a day can you realistically review?
The number of reviews does not depend on how much time you have, but on how many new cards you let in. Every new card then comes back several or a dozen times over the following weeks, so twenty new ones a day adds up, after a month, to several hundred reviews a week. That is not a side effect, it is arithmetic.
A sensible start is ten to twenty new cards a day, and raising that number only once the review queue has held steady for a week or two. In Glimsy the daily new-card limit is set per deck, and it is the one number with which a user genuinely steers the workload.
How do you write a flashcard you can actually learn from?
The rule that matters most: one card, one fact, phrased so that the answer is short and unambiguous. A card that says “discuss the causes of the French Revolution” has no answer you could recall in a few seconds, so every review ends in guessing and a verdict of “more or less”.
The second rule: do not study from cards you do not understand. Flashcards consolidate, they do not explain — understanding has to come first. And the third, most often skipped: ask in both directions only where it genuinely makes sense. With vocabulary, yes; with definitions, usually not, because a definition rarely points back to one single term.
Choosing a flashcard app
There are plenty of flashcard apps and at first glance they look alike: you add cards, then you review them. The difference sits in what you cannot see — whether the program decides for itself when to show you a given card, and what happens once there is a genuinely large amount of material. Below are the questions we hear most often from people who are in the middle of choosing.
What separates an app with a review algorithm from a plain set of flashcards?
A plain set shows cards in the same order or at random, and whether you come back to a hard card is entirely up to you. An app with a review algorithm keeps a separate account for every card: it remembers how the card has been going for you and sets the next meeting so that it lands just before you would forget. The card you struggle with returns tomorrow; the one you have known for months, in six months.
The practical difference shows up with larger material. At thirty cards both routes work the same. At a thousand, deciding by hand what to review today stops being feasible — and that is where the algorithm stops being an extra and becomes the only reason the whole thing does not fall apart.
I am looking for an alternative to Anki — what should I look at?
Three things are worth checking before you move. The first is the algorithm: does the app set the dates itself, and does it do so with a known, documented algorithm — today that means FSRS. The second is the move itself: can it read .apkg or .colpkg files together with images and audio, and do the cards start where they were rather than from scratch. The third is everyday comfort: how it works on a phone, whether reviews run without internet, and how much you have to configure before you see your first card.
Glimsy was built around exactly those three points: FSRS on settings you do not have to tune, an import of both formats that carries your study progress over, and a review session that works offline. We treat Anki as the reference point here, because it is what we studied with ourselves — and that is why the import was one of the first things we built.
Which flashcard app should I use instead of Quizlet?
Start by naming what you are missing — everything else follows from that. If it is long-term studying, meaning material for a whole semester or vocabulary meant to last for years, look for a tool with spaced repetition where the dates are computed by an algorithm rather than by you. If it is your own material, check whether a list can be read in from a file and whether a card can hold an image and audio. If it is studying on the move — whether a session runs without internet.
Glimsy answers the first of those needs above all: the order is governed by FSRS rather than by a card’s place in a set, and that holds equally for decks written by hand, read in from a file, or imported from Anki. The app runs on iOS, Android and in the browser, and the interface speaks English, Polish, Spanish, German and French.
How do you move flashcards from Anki to another app?
Anki exports its data to .apkg files (a single deck) and .colpkg files (a whole collection) — those are the formats every importing tool looks for. Inside sits not only the content of the cards but also images and audio, plus the state of each card: how long its interval is, how hard it has tended to be, and when the next review falls.
Glimsy reads both formats together with their attachments and replays that state through its own FSRS engine, so that a card starts where it already was rather than as a new one. In practice that means the day after the move your reviews fall on the same dates as before, not back at the beginning. The import is a paid-plan feature.
Your own material, and sharing decks
Hardly any deck is built from nothing. Someone has a vocabulary list in a spreadsheet, someone was handed a set by a coursemate, someone wants to add photographs from a lecture or a recording of a pronunciation. These three questions come up most often.
How do you send someone your flashcards?
The usual route is to generate a code or a link that the other person opens on their side — nobody has to be added as a friend and there is no shared account to set up.
In Glimsy a deck gets an eight-character code in the form ABCD-EFGH, a link and a QR code; which one you use depends on whether you are sitting next to each other or typing in a chat. The recipient gets their own copy of the deck: from that moment they study it at their own pace and can change it on their side without touching yours. The code can be made single-use, or revoked later if the deck was only ever meant for one person.
Can you build flashcards from a file, say a vocabulary list in a spreadsheet?
Yes, and it is the fastest route if the material is already sitting in a table somewhere. The usual layout is two columns — question and answer — and the formats that carry such files are CSV, XLSX, JSON and TXT. Glimsy reads all four.
The trap is always the same: a file of sentences pasted straight out of a textbook produces cards you cannot recall, because the question has five correct answers. Ten minutes spent trimming the columns before the import saves weeks of reviewing something that was never going to stick.
Can you put images and audio on a flashcard?
Yes, and for some material that is not decoration but the condition for the card to work at all. Anatomy, art history or pronunciation in a foreign language reduced to a description in words end up testing something other than what they were meant to — recognition takes the place of recall.
Cards in Glimsy are written in a rich text editor, and images (PNG, JPEG, WebP) can be inserted on every plan, the free one included. Attaching your own recordings — MP3, WAV, OGG and their relatives — is a paid-plan feature. Audio files that arrive with an Anki deck land on the cards by the same route.
Studying for an exam
Flashcards reach most people a few weeks before an exam — and that is the moment when it is easiest to do yourself harm, because the method works in the opposite direction to the way people usually study before a deadline. Three questions we get most often on this.
How far ahead of an exam should you start using flashcards?
A practical rule: as many weeks ahead as there are chunks of material to master, and never fewer than three. Spaced repetition needs time by definition — if three days pass between your first and last meeting with a card, that is not spaced studying, it is cramming with an extra step.
The second conclusion matters just as much and is skipped more often: the daily number of new cards is set at the start, because it determines how many reviews land on you in the final week. Letting half the material in four days before the deadline produces a queue that cannot be worked through, and it usually ends with the deck being abandoned.
How do you keep on top of it when there is a huge amount of material?
Three things make the biggest difference here. The first is splitting the material into decks that match how it is examined — one per subject or per topic, not one enormous deck for everything; only then can you see where you stand. The second is setting the daily new-card limit separately for each of them, because that is what decides how many reviews land on you three weeks from now. The third is keeping the cards short: with a lot of material, that is the one thing that holds a session to a reasonable length.
The rest is a matter of regularity, and there no app is going to do the thinking for anyone. In Glimsy we add a garden to that, which grows for reviews done day after day — a small thing, but in practice it works better than a resolution.
Can you study without internet, on a train or a plane?
Reviews, yes — and that is the part that matters, because being on the move is when most of them get done. In Glimsy decks and cards are held locally, and your answers queue up and are sent once the network is back; a session on the underground looks exactly like one at home.
A connection is needed for the things that happen outside studying itself: sharing a deck with someone, the trash, and changes to your account. In practice that means you study normally while travelling and do the housekeeping once you are back.
One last thing
Glimsy is a young app and there is plenty of work still ahead of us. The direction does not change, though: every new thing has to take work off the learner rather than add another setting — which is why, before we add anything, we ask first whether it can be done without putting unnecessary questions to the user.
If something is missing here, or we have described something imprecisely, write to us — one of the two people who build this app reads it.