# NeverCram > AI-powered adaptive learning platform that helps college students study smarter using AI-generated flashcards and spaced repetition (FSRS algorithm). NeverCram transforms any study material into interactive review sessions. Students can generate flashcards from text, PDFs, URLs, and YouTube videos, then review them using an optimized spaced repetition schedule powered by the FSRS-6 algorithm. ## Features - AI-powered flashcard generation from text, PDFs, URLs, and YouTube videos - Spaced repetition scheduling using the FSRS-6 algorithm - Intelligent learning modules with guided study paths - Multiple card types: basic, cloze deletion, and multiple choice - Progress tracking with streaks, XP, and performance analytics ## Links - [Website](https://www.nevercram.app) - [Blog](https://www.nevercram.app/blog) - [Changelog](https://www.nevercram.app/changelog) - [Privacy Policy](https://www.nevercram.app/legal/privacy) - [Terms of Service](https://www.nevercram.app/legal/terms) ## Blog Posts ### FSRS-6 vs SM-2: why the scheduler upgrade matters A plain-English explainer on what FSRS-6 actually does differently from SM-2 — and the retention targets where the delta is largest. # FSRS-6 vs SM-2: why the scheduler upgrade matters If you opened Anki on a new machine in 2026 and did not touch the deck options, you might still be running SM-2. The algorithm has been Anki's default for two decades, it still ships in a lot of third-party tools, and most community decks assume it. FSRS-6 is the upgrade — trained on hundreds of millions of real reviews, shipped as the current Anki default for new profiles, and available as an opt-in for everyone else. The practical question is whether switching is worth the one-time friction. Short answer: yes for anyone studying seriously past a four-week horizon; not really for casual study on a single subject. The long answer is worth understanding, because most comparison articles overstate the delta by citing benchmarks at a retention target that is not necessarily the one you should be using. ## What is FSRS-6 in one paragraph? FSRS stands for Free Spaced Repetition Scheduler. Version 6 shipped in late 2025 and is trained on roughly 700 million reviews contributed by about 20,000 volunteer Anki users ([fsrs4anki wiki, checked 2026-04-18](https://github.com/open-spaced-repetition/fsrs4anki/wiki/abc-of-fsrs)). It models three things per card: stability (how well the memory holds over time), difficulty (how hard the card is for you specifically), and retrievability (your current probability of remembering it). Every review updates the model; the next-due interval is whatever keeps retrievability at your chosen retention target, typically 90%. ## What is SM-2 and why is it still the default anywhere? SM-2 is the SuperMemo-2 algorithm, published by Piotr Wozniak in 1987. It uses a single "ease factor" per card plus a small set of hand-picked multipliers: rate a card "Good," next interval is previous interval × ease factor. Rate it "Hard" and ease factor drops a little. Rate it "Again" and ease factor drops a lot and interval resets. SM-2 is still the default in places for two reasons. First, inertia — two decades of tooling, deck exports, and scheduling assumptions are built around it. Second, simplicity — the algorithm has maybe six parameters total, which makes it trivial to implement and intuitive to debug. FSRS-6 by contrast has seventeen weights that a trained optimizer tunes per user. SM-2 is adequate for casual study. If you review 50 cards a day on one subject for a six-week semester, SM-2 will get you to the exam. The failure mode is not catastrophic — it is gradual. Intervals grow too fast for cards you find hard and too slow for cards you find easy, and the cumulative cost is more reviews than you strictly needed. ## Where does FSRS-6 actually outperform SM-2? The headline benchmark everyone cites: FSRS reduces the number of reviews required at matched retention by 20–30% ([fsrs4anki wiki, checked 2026-04-18](https://github.com/open-spaced-repetition/fsrs4anki/wiki)). That number is real but it is also load-bearing on an assumption most articles skip: the benchmark uses a 90% retention target. Three places the delta is largest in practice: - **Long horizons.** SM-2's ease factor has no memory of why a card was hard — it just notices you struggled. FSRS-6 models stability as a function of prior reviews, so a card you kept failing will be scheduled more conservatively than a card you just happened to miss once on a bad day. The difference compounds over months. For anyone studying toward a one-year horizon (USMLE Step 1, bar exam, comprehensive finals), FSRS's 20–30% review reduction is conservative. - **Large decks.** On a 500-card deck the absolute number of saved reviews is small. On a 20,000-card deck — AnKing, a full language vocabulary, a multi-year knowledge base — those percentages translate into hours per week. Scale is where the algorithm upgrade becomes visible. - **Mixed-difficulty decks.** SM-2 applies the same ease-factor mechanics to every card. FSRS-6 models difficulty separately and schedules the hard cards more often without punishing the easy ones. If your deck has a bimodal distribution — half the cards you nail, half you struggle with — FSRS-6's gain is larger than the average benchmark suggests. ## Where does the benchmark overstate? Here is the contrarian read: the 20–30% figure assumes you are optimising for 90% retention. Push your retention target higher — say 95% for an exam you really cannot miss — and the gap narrows because any scheduler at that target is reviewing you close to the floor of the forgetting curve. Push it lower — 80%, which is defensible for casual self-study — and the gap narrows again because SM-2's imprecision matters less when you are already accepting more forgetting. The sweet spot for FSRS's advantage is 88% to 92% retention, which happens to be where the algorithm was tuned. If you have consciously chosen a different target, the improvement you will actually see is different from the advertised benchmark. Most learners have not consciously chosen a target; they are running whatever the default is. At Anki's 90% default, FSRS-6 is the better algorithm. That framing is accurate in the common case — but you can read "20–30% fewer reviews" into an internal forum post and arrive at an expectation that does not match your own retention preference. Know the target you are actually optimising for before you get excited about the delta. ## When is SM-2 good enough? Named concession: SM-2 is adequate for short-horizon, single-subject, low-stakes study. If you are learning 200 Spanish vocabulary words for a two-month trip, SM-2 gets you there. If you are drilling 80 terms for a Friday quiz, SM-2 gets you there. The algorithm becomes hostile only at scale — when your deck is large, your horizon is long, or your workload is mixed across subjects with different difficulty profiles. The failure mode of staying on SM-2 in the hostile regime is not a disaster. It is a slow tax. You will review more than you needed to, you will forget cards you would have kept if the scheduler had spaced them differently, and you will not know what you lost because the counterfactual is invisible. The argument for upgrading is not that SM-2 is broken; it is that there is a measurable better option and the one-time switching cost is low. ## Does Anki default to FSRS or SM-2 now? As of Anki 23.12 and forward, FSRS-6 is available as the scheduler and is the default for new profiles on recent installs ([Anki manual — deck options, checked 2026-04-18](https://docs.ankiweb.net/deck-options.html)). Existing profiles continue on whatever they had unless you flip the switch in deck options. If you installed Anki three years ago and have never been to deck options since, you are almost certainly still on SM-2. NeverCram ships FSRS-6 as the default for every user with no opt-in required. This is intentional — most learners will not change a default, so whichever scheduler is on at signup is the one they actually run. ## How do you switch, and what breaks? In Anki: Deck options → Advanced → enable FSRS, then optionally run the FSRS optimiser on your review history to tune the seventeen weights to your specific retention pattern. The optimiser takes a few seconds to a minute depending on how large your review log is. Existing scheduling state is translated; intervals will recalibrate over the first week of reviews but no cards are lost. What breaks: nothing in the deck itself. What feels different: due counts may spike or drop in the first few days as FSRS-6 rescores the backlog. This is the algorithm correcting SM-2's cumulative drift, not a bug. Give it a week before judging. ## What this means for your study choices If you are already on FSRS-6 (new Anki install, NeverCram, or any modern scheduler-first tool), you do not need to do anything. If you are on SM-2 and studying at scale or toward a long horizon, flip the switch this week — the one-time friction is small and the cumulative retention advantage compounds. If you are on SM-2 and studying casually, it is a preference call, not a requirement. The deeper shift is that "FSRS vs SM-2" stopped being a live debate once FSRS-6 became the default in the tools that matter. The real comparisons in 2026 are between tools that both run FSRS-6 — authoring workflows, session design, platform support. Picking your scheduler is table stakes. Picking your tool is the interesting decision. [Try NeverCram free →](/signup) --- _Last updated: 2026-04-18. Drafted with AI assistance, edited and fact-checked by Rahul Ragi. Sources verified: [fsrs4anki wiki](https://github.com/open-spaced-repetition/fsrs4anki/wiki), [Anki manual](https://docs.ankiweb.net/deck-options.html)._ --- ### MCAT spaced repetition schedule: 6-month plan A realistic 6-month MCAT spaced repetition plan — deck size, daily cadence, and when the full 20k AnKing deck is actually wrong for you. # MCAT spaced repetition schedule: the 6-month plan that actually fits your life If you are three to six months out from your MCAT test date, the dominant Reddit advice is to download AnKing — roughly 20,000 cards covering every content category the AAMC publishes — and drill it daily until test day. That plan works for a small number of candidates with 12+ months of runway and the disposition to drill for two hours every morning. For the other 80% of test-takers, the full deck is the reason they quit spaced repetition in month two and never rebuild the habit. This guide is a realistic schedule for a 3-to-6-month prep window. It assumes you have a job, a semester, or a life around the studying, and that you would rather retain 4,000 cards than abandon 20,000. The algorithm is not the bottleneck here — deck size and daily cadence are. ## How long does MCAT content actually take to master? The AAMC content outline lists roughly 10 broad categories across the four sections ([AAMC, checked 2026-04-18](https://students-residents.aamc.org/prepare-mcat-exam/mcat-content-outlines)). For a student with solid undergrad coursework in the sciences, the unique-fact count that needs active recall — mechanisms, pathways, named equations, named systems — is closer to 2,500 than 20,000. The AnKing figure is high because it includes variations, reverse cards, and mnemonic duplicates that exist to catch every possible question framing. You do not need all of them. The test is four hours forty-five minutes of timed passages. What it rewards is fast retrieval of a mid-sized core plus applied reasoning. Spaced repetition is optimal for the retrieval half. The applied-reasoning half is earned in practice passages, not flashcards — and no amount of deck grinding substitutes for three-and-a-half passage-based practice exams. ## What is the right deck size for 3-6 months? Cap your active working set at **3,000 to 5,000 cards** for a 3-6 month window. Here is why. Retention math: FSRS at a 90% target schedules a mature card roughly once every 45 days on average ([fsrs4anki wiki, checked 2026-04-18](https://github.com/open-spaced-repetition/fsrs4anki/wiki/abc-of-fsrs)). A 5,000-card mature deck produces about 110 reviews per day at steady state. A 20,000-card deck produces 440. At realistic review speeds of 10-15 seconds per card, 440 reviews is 75-100 minutes before you open a practice passage. Most candidates cannot sustain two hours of pure review plus three hours of practice for six months. They hit week four, fall behind by 200 cards, then 500, then abandon the scheduler entirely. The failure is not algorithmic — it is volume. A 3,000-5,000 card deck produces 65-110 daily reviews. That is 15-25 minutes at steady state. Sustainable. ## What is the right daily cadence? Three blocks, roughly one hour total on non-practice-exam days: - **Block 1 — 15-25 minutes of scheduled reviews.** Whatever the scheduler has due today. Do not review ahead. Do not skip. If due count spikes because you missed a day, let FSRS rebalance on its own over the next 3-4 days rather than cramming the backlog. - **Block 2 — 15-20 minutes of new cards.** Cap new cards at 20-30 per day. At 25/day over six months you add 4,500 cards — which is the working-set ceiling. Hitting the cap by month five is the plan, not a problem. - **Block 3 — 25-35 minutes of passage practice or application.** Spaced repetition drills retrieval. It does not drill reasoning under time pressure. Every study day needs a passage block, even a short one. On full practice exam days (one per week from month three), drop to reviews-only. Half-length full exams run 2h20min; a full UWorld-style exam runs 4h45min. You do not have the bandwidth for new cards on those days and you do not need them. ## How to build the schedule in 10 minutes Use whichever spaced-repetition tool you prefer — the mechanics are the same in any FSRS-6 implementation. Steps: 1. **Pick a content source.** One textbook per section (Kaplan, Princeton Review, or the AAMC official materials). Do not mix three sources in the first pass; you are generating redundant cards and tripling your deck size. 2. **Generate or author the core deck.** For each chapter, extract the named mechanisms, equations, and systems. Target 60-120 cards per textbook chapter. An AI card generator gets you a draft in a minute; a human editor takes another 5-10 to cut compound cards and leading questions. 3. **Triage before you commit to reviewing.** Delete 30-40% of generated cards on first pass. Every card you keep is a review-slot commitment for six months. Definitional cards for terms that appear once in the chapter are net-negative. 4. **Seed the scheduler.** Review every surviving card once on the day you add it. This gives FSRS-6 enough signal to produce realistic intervals by day three. 5. **Set retention target.** 90% is the right default for MCAT. Going higher (95%) means more reviews for marginal retention gain. Going lower (85%) means faster through the deck but more forgotten mechanisms on test day. Total setup time per chapter: 15-20 minutes once the workflow is established. Over a full content-review cycle (8-12 weeks), this is where most of the front-loaded effort sits. Cepeda et al.'s 2006 meta-analysis in _Psychological Bulletin_ found the strongest spaced-practice gains for single-concept items at retention intervals matching the desired testing horizon ([Cepeda et al. 2006, checked 2026-04-18](https://pubmed.ncbi.nlm.nih.gov/16719566/)). For MCAT that means individual mechanism cards, not compound cards, and a retention target aligned with your test date — which FSRS handles automatically once you set 90%. ## The content category that breaks the 3-5k cap CARS (Critical Analysis and Reasoning Skills) is the exception. There is no card-based approach to CARS that reliably outperforms passage practice, because the section tests reasoning on unfamiliar material rather than retrieval of memorised content. Do not build CARS flashcards. Do passage practice daily from month one to test day — 1-2 passages per day under time. This splits your study schedule cleanly: science cards live in the scheduler, CARS lives in timed practice, and the two do not overlap in method even though they overlap in calendar time. ## The last 30 days From day -30 to test day, stop adding new cards. Your working set freezes. Daily cadence: - 30-45 minutes of scheduled reviews — this grows slightly as the scheduler pushes cards for final consolidation - 1-2 full or half-length practice exams per week - 3-5 days of targeted review on sections where practice-exam scores are below your target The temptation in the final month is to panic-add content. Resist it. At this horizon, any new card has not had time to consolidate, so adding 500 cards in week -3 produces 500 cards you will half-remember on test day. Stronger move: drill the existing set harder at 92-95% retention target instead of adding new content. ## Named concession: when the full AnKing deck is the right call For candidates with 12+ months of runway, no concurrent coursework, and genuine enjoyment of daily drilling — the full AnKing deck works. 20,000 cards spread over 14 months produces roughly 45-60 minutes of review per day at steady state, which is sustainable when the rest of your schedule has been cleared to support it. The deck is high-quality, carefully maintained, and represents years of community curation. It is not the right starting point for the person with 4 months, two jobs, and three premed courses still running. The reason to write this guide is that the default advice does not acknowledge which cohort it serves — and the cohort it serves well is smaller than the cohort it gets recommended to. If you are the 12-month candidate, ignore this guide and run AnKing. If you are the 3-6 month candidate, the schedule above is the one that will still be running in week 20 when the other one collapsed in week 8. ## What this looks like on test day A candidate who ran the 4,000-card schedule for six months shows up to the MCAT with ~3,800 cards at 90% mature retention, three-plus practice exams under the belt, and daily-CARS-passage reasoning warm. The scheduler did its job — retrieval is fast, the content base is solid. The remaining variable is how well you read passages under time pressure, and that is the variable practice exams train. That is the honest deliverable of spaced repetition for MCAT: it handles the retrieval half completely and leaves the reasoning half to the method designed for it. No deck, at any size, substitutes for passage practice. [Try NeverCram free →](/signup) --- _Last updated: 2026-04-18. Drafted with AI assistance, edited and fact-checked by Rahul Ragi. Sources verified: [AAMC MCAT content outlines](https://students-residents.aamc.org/prepare-mcat-exam/mcat-content-outlines), [fsrs4anki wiki](https://github.com/open-spaced-repetition/fsrs4anki/wiki/abc-of-fsrs), [Cepeda et al. 2006](https://pubmed.ncbi.nlm.nih.gov/16719566/)._ --- ### How to turn a PDF into flashcards (without the noise) A 10-minute workflow to turn a PDF into flashcards you will actually review — and the cases where no tool can help. # How to turn a PDF into flashcards (without the noise) If you are a med student or self-taught learner sitting on a 600-page PDF the week before exams, the fastest way to turn a PDF into flashcards is to feed it to an AI tool, then delete the bottom third of the cards before you review anything. That triage step is what separates a 15-minute daily review from a two-hour grind. Most tutorials skip it, which is why most people's decks end up abandoned within a week. This guide walks through the real workflow — what to do, what to skip, and when the PDF itself is the problem. ## What does "turn a PDF into flashcards" mean in 2026? Three years ago, this meant copying sentences into a flashcard app by hand. Today, it means handing the PDF to an AI tool that reads it, extracts the claims, and emits question-answer pairs. Generation takes under a minute for a typical textbook chapter. But the tool produces more cards than you need. A 40-page chapter routinely yields 150+ cards out of the box. For retention, you want 40 to 60 well-formed cards per chapter, not 150 middling ones. Quantity is not the deliverable. Retention per hour studied is. ## Why do most PDF-to-flashcard workflows produce too many bad cards? The bottleneck is triage, not generation. Every AI flashcard tool on the market in 2026 — NeverCram included — will generate cards from almost any readable content. What differs between good and bad workflows is how quickly you cull the output. Three patterns show up repeatedly when reviewing generated decks: - **Definitional cards for terms that appear once.** The tool sees a bolded term, makes a card. You never encounter the term again. That card is a forgetting-curve tax. - **Wrapping cards.** "What is X?" answered by "X is Y." Then a second card: "Y is the definition of what?" You are answering the same fact twice. The Anki manual flags this as a common anti-pattern under overlapping tests ([Anki manual — card design, checked 2026-04-18](https://docs.ankiweb.net/editing.html)). - **Compound cards.** "What are the four stages of cellular respiration and what happens in each?" That is four cards in a trench coat. You will forget one stage and the whole card comes back, dragging the other three with it every single review. The fix is the same in all three cases: delete on first pass. Do not try to salvage. ## What makes a flashcard actually worth keeping? After you run the first triage pass, the cards that remain should pass three checks before you commit to reviewing them. First, the question tests one fact. Not "one topic" — one fact. If you can answer a card correctly while getting an adjacent fact wrong, it is atomic. If one wrong sub-fact fails the whole card, split it. Second, the question does not leak the answer. "What organelle produces ATP through oxidative phosphorylation?" leaks less than "What does the mitochondrion do in cellular respiration?" — the second one gives away the subject. Rewrite leading questions or delete them. Third, the answer is something you would recognize as wrong if you said a different answer out loud. Cards whose answers are long and descriptive ("describe the Krebs cycle") are essays, not flashcards. They train recognition of your own prior writing, not recall of the underlying fact. Cards that fail these three checks are not partial credit. They are net-negative: they occupy review slots that would otherwise go to cards that actually build memory. ## How do you get usable flashcards out of a PDF in under 10 minutes? Here is the workflow, tested against a 42-page immunology chapter: 1. **Upload the PDF.** Let the tool generate with default settings. You are testing how the defaults respond to your specific content before you touch any knobs. 2. **Rapid-triage the generated cards.** Keep or delete each card without reading in detail — you are filtering for cards that test something you need to recall, not evaluating correctness yet. Aim to delete 40 to 60 percent on this pass. Anything you hesitate on, delete. 3. **Re-read the survivors slowly.** Now check correctness. Merge compound cards into atomic pairs. Rewrite any card whose question leaks the answer. Apply the three checks from the previous section. 4. **Seed the scheduler.** Answer every surviving card once on day one. This gives the FSRS scheduler enough signal to produce realistic intervals by day three. The fsrs4anki wiki documents the 90 percent retention target FSRS optimizes toward and how it differs from older algorithms like SM-2 ([fsrs4anki wiki, checked 2026-04-18](https://github.com/open-spaced-repetition/fsrs4anki/wiki)). 5. **Trust the schedule.** When the scheduler marks a card as not due, do not review it early. The gap between reviews is the mechanism; collapsing the gap collapses the benefit. End to end on a 42-page chapter: around 8 minutes of generation plus rapid triage, then 15 minutes of careful review. After that, the daily review queue manages itself at 10 to 20 minutes a day. On atomicity: the 2006 Cepeda et al. meta-analysis in _Psychological Bulletin_ found the strongest spaced-practice effects on items that tested a single concept at a time. Compound cards dilute that effect. The triage step is where you enforce atomicity. ## What kind of PDF will defeat any tool? Some PDFs cannot be turned into good flashcards. It is worth naming these before you spend an hour fighting the tool. - **Scanned photocopies without an OCR layer.** If the PDF is a sequence of images, no generation tool can read it. OCR quality caps output quality. Run the file through an OCR pass first or pick a different source. - **Slide decks where the meaning is in figures.** If most of a slide's information is a diagram, extracted text alone will miss the point. Generate cards from the lecture transcript instead. - **Equation-heavy sources with raster equations.** LaTeX-rendered equations often survive extraction; image-rendered ones do not. Test the first page before investing in the workflow. These are caps on the tool, not failures of the tool. The 10-minute workflow above assumes the PDF is native text or well-OCR'd text. ## What the demos don't show Every vendor in this category — NeverCram included — has a demo where a clean PDF becomes a polished deck. That demo is real. It is also the 80th-percentile PDF. On the 40th-percentile PDF, you will do triage work. The honest framing is that the tool gets you from zero to 150 draft cards in a minute, and you get from 150 draft cards to 50 useful cards in eight more. If a vendor implies otherwise, ask to see the second step live. [Try NeverCram free →](/signup) --- _Last updated: 2026-04-18. Drafted with AI assistance, edited and fact-checked by Rahul Ragi. Sources verified: [fsrs4anki wiki](https://github.com/open-spaced-repetition/fsrs4anki/wiki), [Anki manual](https://docs.ankiweb.net/editing.html)._