I Switched From Anki to a New Flashcard App. Here's What Happened to My 400-Day Streak.

I'm going to be honest about something. I'm the person who built Imprimo, so take this with whatever amount of salt you think is appropriate. But the experience I'm about to describe is real, and it's the reason the app exists.

400 days in

I started using Anki seriously during my second year of software engineering. Like most people who discover spaced repetition, I began with a deck someone had shared covering data structures and algorithms. I set up the default settings, started reviewing, and within a week I was hooked. The streak counter ticked up. My exam scores improved. Spaced repetition worked exactly as advertised.

By day 100, I was reviewing about 150 cards daily. Manageable. By day 200, it was pushing 250. By day 300, I was regularly hitting 300+ cards and spending over an hour just on reviews before I could add any new material.

The math was against me. SM-2's ease factors had dropped on hundreds of cards, creating that "ease hell" problem where cards come back way too often. I tried the common fixes: installing the "ResetEZ" add-on, adjusting the starting ease, changing the interval modifier. Each fix helped for a week or two, then the review count crept back up.

By day 400, I was spending 90 minutes every morning on reviews. I'd skip weekends to get a break, then face a wall of 600+ cards on Monday. I dreaded opening the app. My streak was a prison sentence, not an achievement. If that daily volume sounds familiar, the article on how many flashcards to review per day covers how to set a limit before the load takes over.

The breaking point

The thing that finally broke me wasn't the review count. It was a moment during a study session where I realized I was mindlessly hitting "Good" on cards I hadn't actually read. I was going through the motions, maintaining the streak, but not actually engaging with the material. My retention rate on hard cards had dropped below 70%.

I'd become a button-pressing machine. The app was supposed to serve my learning, and instead my learning was serving the app's scheduling demands.

I also hated the UI, and I know that sounds superficial. But when you're spending 90 minutes a day in an app, the experience matters. Anki looks like it was designed in 2006, which it was. The default card templates are plain HTML with no styling. The settings screen has dozens of options that require reading wiki documentation to understand. Every improvement requires hunting through an add-on marketplace of varying quality.

I get why Anki is the way it is. It's open source, built by a small team, and it prioritizes flexibility over polish. That's a valid approach. But I wanted something that felt calmer to use, especially when I was already stressed about exams.

Building something different

I started prototyping Imprimo during a semester break. As a software engineering student, building mobile apps was something I already knew how to do, so the initial version was just a wrapper around FSRS with a nicer card UI. But the more I dug into the problems I'd experienced with existing spaced repetition apps, the more I realized the issues were structural, not cosmetic.

The scheduling was the most obvious issue. SM-2 was the root cause of my review inflation. FSRS solves this directly. It models stability and difficulty separately, so a card that's hard for you doesn't get punished with a permanently low ease factor. In my testing, switching to FSRS reduced daily review counts by about 25% at the same retention target.

Then there was content creation. Half the time I "spent studying" in Anki was actually spent writing cards. Formatting them. Fixing typos. Deciding how to break down a concept into atomic pieces. This is real work that doesn't feel like studying because it isn't. It's authoring. I wanted AI to handle the decomposition: feed it a PDF of lecture notes and get back a set of cards that follow good flashcard principles (one fact per card, clear question, unambiguous answer).

And motivation. Anki's interface is purely functional. There's no sense of progress beyond the streak counter. No visual feedback on retention trends. No feeling of accomplishment after completing a session. I wanted the review experience to feel satisfying, the same way a well-designed workout app makes exercise feel more approachable than a spreadsheet would.

What actually changed

I imported my Anki decks into Imprimo via the .apkg import. FSRS recalculated all the scheduling state, and immediately my daily review count dropped from around 280 to 200. Same cards, same retention target, fewer reviews because the algorithm was more accurate about when I actually needed to see each card.

The bigger difference was psychological. The review sessions felt shorter, partly because they were, and partly because the app was more pleasant to use. The paper-like card design, the haptic feedback on ratings, the little celebration screen after finishing a session: these things sound trivial written down, but they change the emotional texture of the experience. I stopped dreading my morning reviews.

After a month, my retention on hard cards climbed back above 85%. Not because I'd changed my study habits, but because the algorithm was better at showing me struggling cards at the right time instead of drowning me in cards I already knew.

What I changed in the first two weeks

Switching apps didn't magically fix everything on day one. I had to change how I approached the system too.

First, I stopped treating every deck like it deserved equal intensity. In Anki I had built up a habit of saying yes to everything: new deck, new subdeck, new tag, more cards, more material, more "just in case" coverage. The switch forced me to ask a more useful question: what am I actually accountable for remembering this month? That led to a smaller active set of cards and fewer zombie decks hanging around just because they once felt important.

Second, I got more honest about what counts as a good card. If a card required re-reading the whole back before I could grade myself, it was a bad card. If the prompt contained three ideas, it was a bad card. If the answer was technically correct but vague enough to let me bluff my way through, it was a bad card. Better scheduling helps, but it doesn't rescue sloppy prompts. The article on writing flashcards that actually work goes deeper on what separates a useful card from a useless one. This is one reason I'm so interested in AI-assisted card generation done carefully: less typing is nice, but clearer decomposition is the real win.

Third, I stopped worshipping backlog zero. In Anki, finishing everything had become a moral event. If I ended the day with due cards remaining, it felt like failure. That mindset pushed me into long, joyless sessions. After switching, I started thinking in terms of effective sessions instead: focused blocks where I was actually retrieving, not just clearing. Counterintuitively, that made me more consistent because the work felt finite again.

If you want the algorithm-level explanation for why the queue changed so fast, the companion post "FSRS vs SM-2: Why Your Flashcard App Is Using a 40-Year-Old Algorithm" goes deeper into the scheduling side. This article is more about the human side of the switch.

Things I miss about Anki

I should be fair. Anki has capabilities Imprimo doesn't match yet.

The add-on ecosystem is enormous. There are add-ons for image occlusion, heat maps, custom scheduling tweaks, note type templates, and dozens of other things. Imprimo has none of that. What we ship is what you get. For power users who've built elaborate Anki workflows over years, that's a real loss.

Anki's card type system is more flexible. You can create arbitrarily complex note types with multiple fields, custom HTML/CSS, and JavaScript. Imprimo supports basic front/back cards and cloze deletions. That covers 90% of use cases, but the other 10% is a real gap for some people.

And Anki is free. Completely, genuinely free on desktop. The mobile app costs money on iOS but it's a one-time purchase. Imprimo will have a subscription model. For students on tight budgets, that matters.

What surprised me most

The biggest surprise was not the lower review count. I expected that once I moved to FSRS. The surprise was how much mental energy had been tied up in the anticipation of reviews.

Before the switch, a part of my brain was always tracking the queue in the background. If I skipped one day, I could feel the next day's burden immediately. If I added new material, I felt the future cost before I felt the benefit. The system was always asking for a little more attention than I wanted to give it.

Once the schedule got saner, that background pressure eased. I still had to study. I still had hard days. But the app stopped feeling like an adversary. That's a bigger product lesson than it looks. Students churn when the emotional texture of the workflow becomes exhausting, often before a single feature is technically missing.

The other surprise was that losing the streak felt good. I thought I would miss the number. Instead, I noticed how much identity I had attached to it. The streak had become proof that I was disciplined, even on days when the actual learning quality was mediocre. Resetting it forced me to evaluate the thing that mattered more: was I remembering better, with less friction, in a way I could maintain?

That question also connects back to the memory-science piece, "You're Studying Wrong: What 140 Years of Memory Research Actually Says". The research consistently rewards retrieval and spacing, not ritual for its own sake. Streaks can be motivating, but they are not the goal.

Who this is for

Imprimo is for the student who knows spaced repetition works but finds the current tools too painful to stick with. The person who downloaded Anki, used it for three months, and quit because the review count got out of control or the interface wore them down. The person who wants to study from their phone on the train without fighting a desktop-era UI.

If you're an Anki power user with 50,000 cards, custom note types, and a workflow you've refined over years, Imprimo probably isn't for you right now. Stick with what works.

But if you're at the point I was at day 400, drowning in reviews, dreading the app, going through the motions without actually learning, maybe it's time to try something built for how people actually study in 2026.

And if you're not at that point yet, that's useful information too. You may not need to switch immediately. You may simply need cleaner cards, better daily limits, and a more honest study routine. Tools matter, but the right timing matters too. A migration is worth it when it reduces friction you feel every week, not when it just gives you a new dashboard to admire for three days.

My streak reset to zero when I switched. I don't miss it.