AdaptivStudy · the full toolkit
Adaptive algorithms to train your brain: three study modes, a set builder that works with the AI you already use, and the machinery that makes it all adapt.
See what's insideAdaptivStudy is a free study web app that turns notes, textbooks, or any topic into an adaptive study set, then drills that set through three modes built on spaced repetition, retrieval practice, and adaptive testing, entirely on your device.
There is no account and nothing to install or unlock. Your sets, progress, and stats live in your browser's local database, and you can export everything to a backup file whenever you like. Here is what is inside.
Every set can be studied three ways, and each mode keeps its own adaptive state. The mode pages go deep on the learning science behind each one; this is the short version.
Flashcards with a real memory model. Every card carries a 0 to 1 memory-strength score that decides when it comes back: strong cards rest, weak cards return, and a mastered card decays by roughly 8% a day so it resurfaces before you would forget it. Once a set is fully learned, maintenance rounds resurface the weakest cards most. More on Retain and spaced repetition.
Your set's questions run in waves, and a question retires only after you answer it correctly on enough consecutive waves; you choose how strict, 1 of 1, 2 of 2, or 3 of 3. A per-question confidence score from 0 to 100 steers difficulty up or down the Bloom scale, missed questions get a review panel between waves, and every answer is saved so you can stop mid-wave and resume in the exact same spot. More on Recall and active recall.
Exam-style sessions matched to an Elo skill rating from 1.0 to 6.9 and mapped onto Bloom's taxonomy, so climbing means deeper thinking, not just harder facts. You flag every answer Confident or Unsure, which scales the rating move, and the results screen turns your run into a letter grade, a Learning Momentum score from 0 to 100, and your strongest and weakest areas with a missed-question review. More on Reason and adaptive testing.
Build cards by hand across five item types: flashcards, multiple choice, true or false, fill in the blank, and ordering.
Paste text in AQL (Adaptiv Question Language) or AFL (Adaptiv Flashcard Language) and the parser auto-detects the format and previews every card before anything is saved. Blocks it cannot parse are surfaced for rescue, never silently dropped.
A four-step wizard assembles a research-grade prompt from your topic or pasted material, calibrated to a full set: 75 questions standard, 105 comprehensive, or the flashcard count you pick. You run it in the AI you already use, whether that is ChatGPT, Claude, or anything else, then paste the result back to parse and save. Your material never touches a server.
Export everything, or everything except progress, to a versioned JSON file, and import it back later with validation and conflict handling.
A bar chart of your accuracy with Today, 7-day, 30-day, and all-time views, plus a trend arrow over your last hundred answers.
A month calendar of your study days with streak connectors and a pulse on today. Streaks count consecutive study days.
Every mode gets set-level pages: set counts, retired counts, average skill, and a toughest-questions list that expands to the correct answer and its explanation.
The stats hub calls out your strongest set and the one that needs work, once a set has enough answers to judge fairly.
A resume card deep-links straight back into your last in-progress session, so getting started again costs one tap.
Today's questions, cards, and XP at a glance, broken down by mode, with a star on the mode you are strongest in.
Everything you do earns XP through six level tiers, from Neural Spark to Adaptive Genius, and unlocks 12 achievements with bronze, silver, and gold tiers, plus a few secret badges to stumble into.
Five pieces of machinery do the adapting across all three modes. Each mode page covers its piece in depth, with the research behind it; this is the map.
Every question carries a Bloom level, from remembering facts up through analyzing and evaluating. Recall runs its waves inside Bloom-banded levels, and Reason climbs the same ladder as your rating grows, so harder always means deeper thinking, not just obscurity. The full treatment is on the Reason page.
The same rating math that ranks chess players estimates your skill and every question's difficulty on one 1.0 to 6.9 scale, which is how Reason keeps each question near the edge of what you can currently handle.
Being right is not the same as being sure. Recall keeps a 0 to 100 confidence score per question that steers its waves, and Reason scales every rating move by whether you flagged the answer Confident or Unsure, so a lucky guess never reads as knowledge.
Objective-Driven Adaptiv Flashcards, the memory model behind Retain: every card carries a continuous memory strength that climbs as you answer well, decays once the card is mastered, and decides on its own when the card should return. No boxes, no fixed intervals.
Recall runs your questions in waves that shrink as items retire on a streak of correct answers, so every pass pours your effort into whatever is still shaky instead of spreading it evenly.
AdaptivStudy is made by DynamIQ Learning LLC, built so that anyone can study anything without an account or a subscription.
Free, on your device, no account. More questions? See the FAQ.