Guide ·
How to turn a lecture PDF into flashcards
How to turn lecture slides, transes and PDFs into flashcards you'll actually remember, whether you type them yourself or let AI draft them.
A 60-slide lecture can hold 80 to 150 facts worth remembering. Typing those by hand takes two or three hours, and you haven’t studied a single card yet. Most students give up halfway and go back to rereading the slides, which feels productive and mostly isn’t.
Here’s how to get from a PDF to a deck you trust. The method is the same whether you make the cards yourself or let AI draft them.
Pick the pages that matter
Most pages don’t deserve a card. Skip the title slide, the outline, the learning objectives and anything that repeats an earlier slide. Keep definitions, mechanisms, numbers, classifications, and whatever your lecturer said “will come out in the exam.”
In a trans, the summary tables and comparison charts usually pack the most facts per page. Start there.
Write one fact per card
The most common mistake is pasting a whole slide onto one card. You end up rereading instead of recalling, and when you get it wrong you can’t tell which part you forgot.
Split it up. A slide on the stages of wound healing turns into several cards:
- Q: What is the first stage of wound healing? A: Hemostasis.
- Q: Which cells dominate the inflammatory stage? A: Neutrophils, then macrophages.
- Q: About how long does the proliferative stage last? A: Roughly days 4 to 21.
Each card asks for one thing. Miss it and you know exactly what to review.
Ask questions you can’t guess
A good card makes you come up with the answer. Recognizing it isn’t enough.
- Ask what, which, why or how. Yes-or-no questions are a coin flip.
- Watch for wording that gives the answer away.
- Don’t ask for “all seven” of anything. Ask for one item, or the item in a specific position.
- If the answer needs a paragraph, you probably have two or three cards.
Write the page number on every card
A week from now some card will look wrong, and you’ll want to check it against the slide in seconds instead of scrolling through 60 pages. So note the page as you go. With AI-drafted cards this matters even more, because checking against your own material is the only way to catch mistakes.
Study only what’s due
Making the deck is maybe half the job. The rest is reviewing each card on the right day. Spaced repetition brings a card back right before you’d forget it, often at first and then days and weeks apart. Modern Anki uses a scheduler called FSRS, and in its published benchmarks it needs fewer reviews than Anki’s older algorithm to keep the same amount in your head.
Twenty minutes of due cards a day beats two hours of rereading the night before. Cramming works for tomorrow’s quiz and not much else.
Letting AI do the drafting
The first four steps are where the hours go, and AI is good at drafting them. The problem is trust. AI can misread a slide, or confidently state something your lecturer never said. Whatever tool you use, make sure you can check each card against the page it came from. If you can’t, you’re studying guesses.
We built Lara around that check. You import a lecture PDF, slides, a trans, photos of handwritten notes or pasted text, and pick the pages. Lara drafts one-fact cards in seconds, and each one shows its page number and a quote from that page. Tap a card to see the page. If Lara can’t find the quote on the page, the card gets a ⚠ flag so you look at it first. You keep, edit or toss each draft, and the ones you keep go on an FSRS schedule. Studying works offline.
Scanned PDFs and photos work too, with one catch. Lara reads the image of each page, but there’s no text to search, so those cards can’t be flagged. Check them against the page yourself.
The first 25 pages are free, which comes to about 70 cards. Studying, editing and writing your own cards stay free forever. If you need more pages each month, see pricing.

Skip the typing.
Drop your lecture into Lara and get cards tagged with the page they came from. Your first 25 pages are free.
Android is in testing. Email support@uselara.app to join.