Skip to content
The Chat Archive
gemini-2.5-pro· Google LMArena Human Preference 140K

answer again: What is the currently best spaced repetition algorithm?

  1. User

    answer again: What is the currently best spaced repetition algorithm?

  2. Assistant

    Excellent question. The landscape has changed significantly in the last couple of years. While there's no single "best" for everyone, the consensus among serious users has shifted.

    Here’s a breakdown of the top spaced repetition algorithms, with a clear winner for most people.

    The Top Contenders

    AlgorithmKey FeatureBasisProsCons
    1. FSRS (Free Spaced Repetition Scheduler)Personalized & AdaptiveMachine Learning (Logistic Regression)• Highly Personalized: Creates a unique memory model based on your review history. • More Efficient: Often results in fewer reviews than older algorithms for the same retention rate. • Open Source & Actively Developed: Continuously improving. • Now the default in Anki.• Requires an initial data set (a few hundred reviews) to become effective. • The underlying math is more complex for users to grasp.
    2. SuperMemo SM-18The Gold Standard of ComplexityData-driven algorithm from decades of research• Extremely Sophisticated: Models dozens of variables related to memory. • Theoretically the Most Accurate: Based on the largest private dataset of learning data in the world. • Developed by the inventor of the field, Piotr Woźniak.• Proprietary & Closed-Source: Only available in the SuperMemo software. • The software has a notoriously steep learning curve. • Not personalized to the individual user in the same way as FSRS.
    3. Duolingo's "Birdbrain" (HLR)Optimized for EngagementMachine Learning (Half-Life Regression)• Massively Scaled: Trained on data from hundreds of millions of users. • Excellent for Language: Highly tuned for vocabulary and short phrases. • Balances learning with keeping users engaged on the app.• A "Black Box": Not transparent or available outside of Duolingo. • Optimized for Duolingo's business goals (engagement) as much as for learning efficiency.
    4. Anki's Legacy Algorithm (based on SM-2)Simple & RobustAn adaptation of SuperMemo's 1990 algorithm• Simple & Intuitive: Easy to understand the logic. • Reliable: It has been proven effective for millions of users over 15+ years. • Doesn't require any initial training data.• One-Size-Fits-All: Uses the same model for every user and every card. • Less Efficient: Generally leads to more reviews than FSRS or modern SuperMemo algorithms.

    The Verdict: What is the Best Algorithm Right Now?

    For the vast majority of people, the best and most accessible spaced repetition algorithm is FSRS (Free Spaced Repetition Scheduler).

    Here's why:

    1. The Best of Both Worlds: FSRS combines the data-driven sophistication of modern algorithms with the open-source, user-friendly ecosystem of Anki. It learns your personal rate of forgetting and schedules reviews accordingly.
    2. Proven Efficiency: Users consistently report that after switching to FSRS, their daily review load decreases while maintaining or even improving their retention rate. This means less time spent on reviews for the same result.
    3. Accessibility: FSRS is now the default scheduler in the latest versions of Anki across all platforms (desktop, mobile). It's easy to enable and works automatically in the background.

    In summary:

    • For the ultimate power user and researcher: SuperMemo SM-18 remains the most complex and theoretically robust algorithm, but it's locked into a difficult-to-use software.
    • For almost everyone else: Anki with FSRS is the clear winner. It offers state-of-the-art, personalized scheduling in a free, open-source, and widely supported platform. It represents the biggest leap forward in practical spaced repetition in over a decade.

Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.