Jump to main content

Generate Customized Study Flashcards with Artificial Intelligence

Abstract

Have you ever needed to make a lot of flashcards for a class, but didn’t have the time–or weren’t sure how to make ones that actually help you learn? Maybe you like studying on the go, or you’re trying to squeeze in a quick 10-minute review session between other things. In this project, you will learn how to use ChatGPT (or another AI tool of your choice like Gemini, Copilot, etc.) to generate effective study flashcards quickly and customize them to fit your class and your study style. You will also experiment with different flashcard formats and prompting strategies to evaluate which types of flashcards are the most accurate, useful, and effective for studying. 

Summary

Areas of Science
Difficulty
Method
Time Required
Very Short (≤ 1 day)
Prerequisites

None

Material Availability

Readily available

Cost
Very Low (under $20)
Safety

No issues

Credits
Science Buddies is committed to creating content authored by scientists and educators. Learn more about our process and how we use AI.

Objective

To use ChatGPT (or another LLM) to create and improve study flashcards from class materials and evaluate how different prompts affect their quality, clarity, and usefulness. 

Introduction

Flashcards have been a study staple for decades. A flashcard is a small study prompt–traditionally a physical card,  but often digital today–with a question, term, or cue on one side and the answer or explanation on the other. They work because they force you to perform active recall. Instead of passively rereading notes or highlighting a textbook, flashcards force you to pull an answer from memory. That little bit of mental effort is exactly what helps information stick.

Two small square flashcards. The left card asks, "What is an LLM?" The right card shows the answer: "An LLM (Large Language Model) is a type of AI (Artificial Intelligence) that is trained on lots of texts so it can understand and generate language.Image Credit: Science Buddies
Figure 1. A card with a question on one side and the answer on the other side.

Flashcards also pair naturally with spaced repetition, a study strategy in which you review material multiple times with increasing intervals between reviews. The goal is to practice before you forget–strengthening memory each time you successfully recall the information. Quick review sessions spread across a day (or across several days) create spacing automatically. Also, because flashcards give fast feedback–you either know it or you don’t–they make it easy to spot which topics you struggle with and revisit those more often than the ones you already understand.

Another reason flashcards work so well is that they fit into real life. You can study almost anywhere: on the train, waiting in line, between classes, or during a short break. This kind of microlearning is powerful because it lowers the barrier to getting started. It’s much easier to commit to “just 10 minutes” than to plan a long study session, which can reduce procrastination and make studying feel more manageable.

So why don’t people use flashcards consistently? A big reason is the upfront work. Writing good flashcards takes time–and making effective ones (clear questions, precise answers, and helpful examples) can be tricky.

That’s where Artificial Intelligence (AI) can help. AI is a broad term for computer systems that perform tasks that normally require human intelligence, like understanding language or generating text. One major type of AI used today is a large language model (LLM)--a system trained on huge amounts of text to recognize patterns in language and generate human-like responses. ChatGPT is an example of an LLM. It can take your notes, vocabulary lists, textbook sections, lecture transcripts, or review guides and turn them into flashcards in seconds.

In this project, you will experiment with how well ChatGPT (or another LLM) can help you create custom flashcards tailored to your class and learning goals. You will prompt ChatGPT to generate different flashcard styles and then evaluate the cards for accuracy, clarity, difficulty, and usefulness. Based on your results, you will revise your prompts and improve the flashcards through multiple rounds of testing and refinement. By the end, you will have both a personalized set of flashcards and a repeatable process for evaluating and improving AI-generated study tools. 

Terms and Concepts

Questions

Bibliography

To learn more about prompting the LLM:

Materials and Equipment

Experimental Procedure

This project follows the Engineering Design Process. Confirm with your teacher if this is acceptable for your project, and review the steps before you begin.

Setting Up ChatGPT

  1. You will need a ChatGPT account. If you do not have one, make one when prompted. You could also use a different LLM of your choice. 

Creating Flashcards with ChatGPT

  1. Start a new chat in ChatGPT.
  2. Choose your source material. You can copy and paste your notes, attach a PDF or document, or even attach a picture of handwritten notes! You can make flashcards from almost anything, such as:
    1. Class notes
    2. Lecture video transcripts
    3. Textbook sections
    4. Websites or online articles (use caution–online sources can be inaccurate or oversimplified)
  3. Decide what kind of flashcards you want. Before you prompt ChatGPT, think about:
    1. How many cards you want
      1. Fewer cards → more general, big-picture concepts
      2. More cards → more detailed coverage
    2. Card format (pick one or mix multiple)
      1. Definition/vocabulary
      2. Short answer
      3. Fill-in-the-blank
      4. Multiple choice
    3. Card variety
      1. Concept questions, definitions, examples, comparisons, “why/how” questions, etc.
    4. Length and style
      1. Short questions + longer explanations, or quick Q/A for fast review
    5. Any other preferences (difficulty level, key terms to include, topics to avoid, etc.).
  4. Decide how you want to save or export your cards. If you are making physical flashcards, you can ask ChatGPT to generate the questions and answers with no specific formatting–then simply copy them from the chat into your index cards. If you are using a flashcard app like Anki or Quizlet, you can ask ChatGPT to format the cards for quick import.
  5. Write a clear, detailed prompt. The more specific you are, the more useful and customized your cards will be. For example:
    1. “Create 20 flashcards using the lecture transcript I uploaded. Make each question 1-2 sentences, and each answer 2-4 sentences. Focus mostly on conceptual understanding, but include a few definition and example cards. Keep the language slightly simpler than the transcript. Format the output as a ready-to-import Anki TSV file (one card per line, with the front and back separated by a tab).”

Improving the Flashcards

Once your cards are imported and you’ve tested a few, take a few minutes to make them better for learning–not just correct.

  1. Once your cards are created, do a quick “test run” to find what needs fixing. Check the criteria you set for your flashcards back in step 3 of the Creating Flashcards with ChatGPT section.
    1. Study or preview 5-10 cards.
    2. Take notes on what went wrong (e.g., too wordy, too easy, unclear questions, missing examples, lots of duplicates, not enough conceptual questions, inaccurate info).
  2. Update your prompt with specific fixes. Copy your original prompt and add a short “revision checklist” based on what you noticed. Common upgrades:
    1. Make cards clearer:
      1. “Rewrite any vague questions so the front is specific and unambiguous.”
      2. “Avoid pronouns like ‘it/this’--repeat the key term.”
    2. Make cards better for active recall:
      1. “Avoid yes/no questions.”
      2. “Use more why/how/compare questions rather than only definitions.”
    3. Control difficulty:
      1. “Make the questions slightly harder by asking for reasoning or application.”
      2. “Simplify the language.”
    4. Improve structure:
      1. “One concept per card. Split multi-part cards into separate cards.”
      2. “Use bullet points for lists on the back.”
    5. Improve coverage:
      1. “Include 60% concept/why/how, 25% definitions, 15% examples.”
      2. “Make sure you cover these topics: ___, ___, ___.”
    6. Reduce redundancy:
      1. “Do not repeat the same fact on multiple cards. Each card should test something distinct.”
    7. Add examples (without making answers too long):
      1. “Include one short real-world example in the answer when helpful (1 sentence max).”
  3. Regenerate and re-import (replace the old deck).
    1. Ask ChatGPT to make the needed changes and create new flashcards.
    2. Preview a few cards again to confirm improvements.
  4. Continue with steps 1-3 until your flashcards meet all of your criteria. Below is an example table you can use to keep track of the changes you made to your flashcards:
    Swipe left to see more
    Table 1. Example data table. You can add more rows to the table for additional revisions. 
    Generation Prompt Notes
    V1 Create 20 flashcards using the lecture transcript I uploaded. The flashcards were too wordy.
    V2 Remake the flashcards to be more concise.  More flashcards with specific examples from class would be helpful. 
    ... ... ...
icon scientific method

Ask an Expert

Do you have specific questions about your science project? Our team of volunteer scientists can help. Our Experts won't do the work for you, but they will make suggestions, offer guidance, and help you troubleshoot.

Variations

  • You can import your flashcards onto an online application like Anki or Quizlet
  • Compare flashcards generated by different LLMs (e.g., Gemini, CoPilot, Claude). 

Careers

If you like this project, you might enjoy exploring these related careers:

Career Profile
Many aspects of peoples' daily lives can be summarized using data, from what is the most popular new video game to where people like to go for a summer vacation. Data scientists (sometimes called data analysts) are experts at organizing and analyzing large sets of data (often called "big data"). By doing this, data scientists make conclusions that help other people or companies. For example, data scientists could help a video game company make a more profitable video game based on players'… Read more
Career Profile
Computers are essential tools in the modern world, handling everything from traffic control, car welding, movie animation, shipping, aircraft design, and social networking to book publishing, business management, music mixing, health care, agriculture, and online shopping. Computer programmers are the people who write the instructions that tell computers what to do. Read more

News Feed on This Topic

 
, ,

Cite This Page

General citation information is provided here. Be sure to check the formatting, including capitalization, for the method you are using and update your citation, as needed.

MLA Style

Ngo, Tracey. "Generate Customized Study Flashcards with Artificial Intelligence." Science Buddies, 30 Mar. 2026, https://www.sciencebuddies.org/science-fair-projects/project-ideas/ArtificialIntelligence_p036/artificial-intelligence/flashcard_llm. Accessed 5 June 2026.

APA Style

Ngo, T. (2026, March 30). Generate Customized Study Flashcards with Artificial Intelligence. Retrieved from https://www.sciencebuddies.org/science-fair-projects/project-ideas/ArtificialIntelligence_p036/artificial-intelligence/flashcard_llm


Last edit date: 2026-03-30
Top
Free science fair projects.