Leitner System vs Anki: Manual Boxes vs SM-2 Software

If you've been spaced-repetition-curious for more than a week, you've hit this exact question: is Anki just the Leitner system on a screen? It's a fair question — they look similar from a distance, and people use the terms almost interchangeably in study forums. They are not the same thing, and the difference is the whole point of this post.

TL;DR The Leitner system is a manual spaced-repetition method: physical boxes, fixed promotion/demotion rules, pass-fail grading. Anki is software that runs the SM-2 algorithm — a per-card, adaptive descendant of the same spaced-repetition idea. Same lineage, different precision. Leitner is great for small decks, beginners, and people who want zero screens. Anki scales to thousands of cards and syncs everywhere, but it's manual to author and has a steep learning curve. If you want SM-2 scheduling without the box-shuffling or Anki's setup friction, that's the gap SmartRecall fills.

Is Anki just the Leitner system? (the direct answer)

No — but they share a bloodline.

Both descend from the same insight: review a card right before you're about to forget it, and push that next review further out each time you succeed. That's spaced repetition in one sentence. Sebastian Leitner described a manual version of it in 1972; Piotr Wozniak formalized a mathematical version (SM-2) in 1987; Anki shipped a software implementation of SM-2 in 2006.

The difference is granularity:

• Leitner moves a card between five (or however many) boxes. Every card in a box gets the same interval. There is no per-card memory of how hard a specific card was — only which box it's currently in.
• Anki's SM-2 gives every individual card its own "ease factor" and its own interval. A card you find trivially easy stretches to months between reviews; a card you keep missing gets reviewed far more often. The schedule is computed per card, every single time you grade it.

So when someone says "Anki is basically the Leitner system," they're right about the idea and wrong about the mechanism. Leitner is fixed-interval buckets. Anki is adaptive per-card math. (If you want the full algorithm-by-algorithm breakdown, I wrote one in SM-2 vs FSRS vs Leitner vs Anki.)

How each one actually works

The Leitner system (manual)

Five boxes. New cards start in Box 1. Get a card right, it moves up one box; get it wrong, it goes back to Box 1. You review Box 1 daily, Box 2 every couple of days, and so on, roughly doubling. The "algorithm" lives in your hands and a stack of index cards. I covered the full setup in the Leitner system explained, so I won't repeat it here — the key property for this comparison is that the intervals are fixed and bucket-wide.

Anki (SM-2 software)

You author a card, and after each review you press one of four buttons — Again, Hard, Good, Easy. Anki uses your answer to update that card's ease factor and compute its next interval. Miss a card and it lapses back into short intervals; nail it repeatedly and it disappears for weeks or months. The scheduling is invisible and automatic — but you still have to type every card yourself.

The comparison table

Honest pros and cons

Leitner — pros

• Tactile and zero-tech. No account, no sync conflicts, no app update breaking your scheduler.
• Forces focus. A stack of index cards can't notify you into a doomscroll.
• The act of sorting is itself a retrieval event. Deciding "this goes in Box 3" is a tiny metacognitive review on top of the card content.
• Perfect for small, visual decks — anatomy diagrams, a new alphabet, the 100 core facts you need burned in.

Leitner — cons

• Doesn't scale. Past a few hundred cards you're shuffling shoeboxes for 30 minutes a day.
• No per-card precision. The card you almost forgot and the card you aced in half a second advance identically.
• No sync, no stats, no audio. Paper can't ride your commute or play a pronunciation clip.

Anki — pros

• Scales to enormous decks and runs the same on phone and laptop.
• Per-card adaptive scheduling genuinely saves reviews on mature decks.
• The most battle-tested SR engine on earth — decades of edge cases shaken out, plus optional FSRS if you want the newer model.

Anki — cons

• Steep learning curve. Deck options, note types, intervals, lapses — most people bounce off the configuration before they ever build a habit.
• You author everything by hand. That's the real wall. Typing 2,000 good cards is days of work most people don't have.
• Easy to over-tune and never actually study.

Leitner vs "spaced repetition software" — clearing up the terms

A lot of searches phrase this as "Leitner vs spaced repetition." That framing is slightly off: Leitner is spaced repetition — it's just the manual, low-resolution version. SM-2 software like Anki is also spaced repetition — the adaptive, high-resolution version. The real axis isn't "Leitner vs spaced repetition," it's "manual fixed-interval spaced repetition vs algorithmic adaptive spaced repetition." Both beat cramming by a wide margin. Both work. They differ in precision and in how much manual labor they ask of you.

And whichever you pick, the underlying study technique that makes either one work is retrieval practice — pulling the answer out of your head rather than re-reading it. If that distinction is fuzzy, active recall vs passive review is the foundational read.

Where SmartRecall fits

Here's the thing both Leitner and Anki have in common: you make every card by hand. Leitner makes you shuffle boxes; Anki makes you fight a config screen; both make you type. For someone staring at a 700-page textbook eight weeks before an exam, that authoring step is the actual bottleneck — not which scheduler picks the next card.

That's the gap I built SmartRecall for. It does three things differently:

• AI generates the cards for you — paste your notes or upload a chapter and our PDF-to-flashcards pipeline turns it into well-formed cards in minutes, so you're not retyping a textbook.
• SM-2 spaced repetition runs automatically — same adaptive, per-card scheduling lineage as Anki, no box-shuffling and no deck-options rabbit hole.
• There's a usable free tier — enough credits to test it on a real chapter before you commit.

I'm not going to tell you it replaces a 20,000-card Anki deck someone has lovingly tuned over three years — it doesn't, and if that's you, stay where you are. SmartRecall is for the people who never get to that deck because authoring it by hand was too much. (If you are weighing the two head-to-head, SmartRecall vs Anki is the honest side-by-side.)

So which should you use?

• Pick Leitner if your deck is small, your material is visual or tactile, and you want to study without a screen. Five boxes, an afternoon, done.
• Pick Anki if you have thousands of cards, a multi-year horizon, and the patience to author and configure. It scales further than anything else and it's free.
• Pick SmartRecall if the part that's actually stopping you is making the cards — and you'd rather have AI generate them and let SM-2 schedule them automatically.

The best spaced-repetition system is the one you'll actually run tomorrow. Leitner's boxes have lasted 50+ years because they work. Anki's SM-2 has powered billions of reviews because it works. And if both stall out at "but I never built the deck," that's exactly the problem worth solving first.

Try generating one chapter's worth of cards and see whether they're the cards you would have written if you'd had the time. That's usually the moment the whole question answers itself.

— Alex