Creating flashcards from lecture slides: step by step, with and without AI

In short: Good flashcards from lecture slides are created in three steps. Identify the lecture's learning objectives, formulate questions for each objective that you have to answer actively, and review the cards over several days at spaced intervals. Copying slide text onto cards does little. Answering questions does a lot. AI can take care of the wording for you, but not the checking and reviewing.
Why flashcards from slides work
Flashcards force you to retrieve knowledge from memory rather than merely recognise it. That is exactly what makes the difference.
In a well-known experiment by Roediger and Karpicke (2006), students read an informational text. Afterwards, they either read it again or wrote down what they remembered from memory. Five minutes later, rereading was ahead: 81% versus 75% of the content remembered. A week later, the result had reversed: 56% after retrieval, compared with just 42% after rereading.
A major review by Dunlosky and colleagues (2013) compared ten common learning techniques. Only two received the highest rating: self-testing and spaced repetition. Highlighting, summarising, and rereading received low ratings. Flashcards combine both strong techniques when you build them properly.
Step by step: flashcards from your slides
1. Find the learning objectives before writing your first card
Many lectures state the learning objectives at the beginning or end, often phrased as “After this lecture, you will be able to…” If they are missing, the slide headings and questions from past exam papers can help. Write down 3 to 8 learning objectives per lecture. Every card you create should contribute to one of them. Everything else is peripheral.
2. One card, one question
A card tests exactly one thing. “Explain the citric acid cycle” is not a flashcard; it is an exam question. Break it into individual questions: Where does it take place? What goes in? What comes out? Which step is rate-limiting?
3. Questions instead of keywords
Write a real question on the front that requires an active answer. Put the shortest complete answer on the back.
| Weak card | Strong card |
|---|---|
| Front: “Pyruvate dehydrogenase” | Front: “Which enzyme links glycolysis and the citric acid cycle, and why is this step irreversible?” |
| Front: “Insulin – effect” | Front: “Name two ways in which insulin lowers blood glucose.” |
| Back: the entire slide text copied | Back: one to three sentences in your own words |
| Fill-in-the-blank text with three blanks | One blank per card |
4. Your own notes belong on the card
What you wrote in the margins during the lecture is often more important than the slide itself: the mnemonic, the professor's example, the question mark beside something you did not understand. Include your memory aids on the card. Turn every question mark into a separate card that you clarify before the exam.
5. Space out your reviews
Do not study a new card ten times in a row. Instead, review it after one day, then after a few days, then after one to two weeks. Cards you know confidently move to a later review. Cards you did not know come back sooner. This principle is called spaced repetition, and it is built into most flashcard apps.
6. Before the exam: plan backwards
Count how many cards you have and how many days remain until the exam. New cards need several reviews. If you start with 400 new cards two days beforehand, you will not get through them in time. Start each lecture's cards in the week the lecture was given.
Creating flashcards with AI: what helps and what to watch out for
AI can formulate questions and answers from lecture slides in minutes. This saves most of the work involved in creating them. Three things still remain your responsibility.
Checking. AI can make mistakes, especially with abbreviations, formulas, and illustrations. Check each card against the slide once before studying it. A wrong card you review ten times is hard to unlearn.
Filtering. AI tends to generate too many cards rather than too few. Keep the ones that contribute to your learning objectives. Delete cards about details that will not matter in the exam.
Reviewing. Finished cards are not a learning achievement. Only spaced retrieval ensures that you still know them in the exam.
Am I allowed to upload lecture slides to an AI tool?
You are generally allowed to use slides your university provides for your studies for your own learning. However, the rights to the slides belong to the instructors. So do not publish either the slides or cards that reproduce their content. Some universities have their own rules for AI tools. If in doubt, check your university's guidance. Also pay attention to where a tool processes your data and whether it uses the data for training.
The most common mistakes
- Copying slides instead of asking questions. Then you are reviewing text rather than retrieving knowledge.
- Too much on one card. If you only half know the answer, you cannot honestly assess whether you know it.
- Only starting at the end of the semester. Spaced repetition needs time that is no longer available two weeks before the exam.
- Never filtering cards out. If you carry around 2,000 cards, 500 of which are unimportant, you lose time for the important ones.
- Ignoring images. Many exam questions refer to illustrations. Create cards that require you to name parts of an illustration.
How Learngine does it
Learngine is a study app that brings together exactly these steps. You upload the slides from a lecture, including your handwritten notes if you like. Learngine identifies the learning objectives, turns them into topics, and generates flashcards, a quiz, and further exercises. Your memory aids stay on the cards. Lumi schedules reviews leading up to the exam and tells you every day what is up next in 10, 25, or 40 minutes. If your note and the slide contradict each other, Coachi works through it with you.
Frequently asked questions
How many flashcards per lecture make sense?
As a rule of thumb: the number of learning objectives multiplied by two to five questions. For a 90-minute lecture with five learning objectives, that is roughly 10 to 25 cards. More only makes sense if the exam tests many details.
How often should I review a flashcard?
A tried-and-tested schedule is: the next day, after about three days, after a week, and after two weeks. Reset cards you did not know to the beginning. Apps with spaced repetition handle this scheduling for you.
Digital flashcards or paper?
Both work if you ask questions and space out your reviews. Digital cards have the advantage that the app manages the review schedule and you can study on the go. Paper helps some people formulate their cards because writing is slower.
Can I rely on AI-generated flashcards?
Not blindly. Use them as a draft and check them against your slides and required exam reading. Numbers, formulas, and technical terms in particular deserve a second look.
Sources
- Roediger, H. L. & Karpicke, J. D. (2006): Test-Enhanced Learning: Taking Memory Tests Improves Long-Term Retention. Psychological Science, 17(3). PDF
- Dunlosky, J., Rawson, K. A., Marsh, E. J., Nathan, M. J. & Willingham, D. T. (2013): Improving Students' Learning With Effective Learning Techniques. Psychological Science in the Public Interest, 14(1). APS summary
Learngine. As of November 6, 2026.
More from Knowledge
All articles on study methods and guides for university and school