FSRS Algorithm Explained: The Modern Spaced Repetition Scheduler

For decades, spaced repetition apps ran on the SM-2 algorithm — the 1987 formula behind SuperMemo and the classic Anki scheduler. It works, but it is dated. The modern standard is FSRS, the Free Spaced Repetition Scheduler, and tegaru now uses it to schedule every flashcard review. This guide explains what FSRS is, how it differs from SM-2, and why it produces fewer, better-timed reviews.

What is FSRS?

FSRS stands for Free Spaced Repetition Scheduler. It is an open-source algorithm, developed by Jarrett Ye and the open-source community, that predicts the state of your memory and schedules each review for the exact moment you are about to forget. It became Anki's built-in scheduler in 2023 and is now the de facto modern standard for serious spaced repetition.

Instead of a single "ease factor," FSRS models your memory of every card with two variables:

• Stability — how long the memory will last. A stability of 30 means that in 30 days your probability of recall has fallen to about 90%.
• Difficulty — how hard the card is for you, on a roughly 1–10 scale. Harder cards gain stability more slowly.

The three pillars of FSRS

FSRS is built on a model of memory called the DSR model — Difficulty, Stability, and Retrievability.

1. Retrievability

Retrievability is the probability that you can recall a card right now. It follows a forgetting curve that decays from 100% over time, and the rate of decay depends on the card's stability. FSRS schedules a review when your retrievability drops to a target you choose — by default around 90%.

2. Stability

Every time you successfully recall a card, its stability increases, so the next interval is longer. The key insight FSRS captures that SM-2 misses: how much stability grows depends on when you reviewed. Reviewing a card the moment it was learned barely strengthens it; reviewing it just as you were about to forget strengthens it a lot. This is the "spacing effect," modeled mathematically.

3. Difficulty

Difficulty adjusts based on how you rate a card. Cards you repeatedly find hard accumulate higher difficulty and grow their intervals more slowly, so you see them more often — exactly where your study time should go.

FSRS vs SM-2: what actually changed

SM-2 uses an "ease factor" (starting at 2.5) and multiplies your interval by it on every success. It has well-known weaknesses: fixed first intervals, no real concept of memory strength, and the dreaded "ease hell," where a few hard ratings permanently tank a card's schedule.

FSRS replaces all of that with a memory model fit to real review data:

• It targets a retention rate, not a multiplier. You can ask for 90% retention and FSRS schedules to hit it.
• It uses the timing of your reviews, not just the rating, to update memory.
• It avoids ease hell because difficulty and stability are separate, bounded quantities.
• It needs fewer reviews for the same retention — studies and large-scale Anki data consistently show FSRS scheduling fewer cards to reach the same recall.

In short: SM-2 asks "how easy was this card?" FSRS asks "how is your memory of this card actually decaying, and when should I catch it?"

How tegaru uses FSRS

tegaru schedules reviews with FSRS-6, the latest version of the algorithm, through the maintained open-source ts-fsrs library — the same core that powers modern Anki.

• Every card stores its own stability and difficulty, updated on each review.
• You rate cards with the familiar four buttons — Again, Hard, Good, Easy — and FSRS converts those into memory updates.
• New cards move through short learning steps (minutes) before graduating to day-scale intervals.
• Reviews are scheduled by calendar day, so a card studied at 11pm is not stuck coming due at 11pm forever.

You do not have to configure anything. Upload a document, generate a deck, and FSRS quietly schedules each card for the moment it will do the most good.

Does the difference matter for you?

If you study a few dozen cards, SM-2 and FSRS feel similar. The gap grows with your deck:

• Big decks (hundreds to thousands of cards) — FSRS meaningfully reduces daily review load for the same retention.
• Exam prep — targeting a retention rate lets you trade a little extra review now for confidence on test day.
• Long-term knowledge — FSRS's longer, better-justified intervals are ideal for languages, medical knowledge, and anything you want to keep for years.

Frequently asked questions

Is FSRS better than SM-2? For almost everyone, yes — it reaches the same retention with fewer reviews and avoids SM-2's failure modes. SM-2 is still a reasonable, simple baseline.

Do I need to understand the math to use it? No. Rate your cards honestly with Again/Hard/Good/Easy and FSRS does the rest.

What happened to my existing cards? tegaru migrated existing decks gracefully — older cards re-initialize into a valid FSRS memory state on their next review, keeping their place in your schedule.

Start studying with FSRS

Every deck you create in tegaru is scheduled with FSRS automatically. Upload your notes, generate flashcards, and let the modern spaced repetition algorithm handle the timing — so you study fewer cards and remember more.