AI flashcard generator
Turn a PDF, document or photo into flashcards you can edit
Upload a PDF, a Word file or a picture of a page. Memor More writes question-and-answer cards from what is in that document, and puts them in a preview you can rewrite, cut or regenerate before the deck is saved.
In our test on 7 September 2026, a one-page lecture handout of just under 200 words produced 17 cards in about 24 seconds. Every one of those cards is reproduced further down this page, unedited.
Get the app and generate a deckFree · iPhone, iPad & Mac · 3 AI generations per week on the free tier
What you can feed it
These are the formats the current app and API accept, checked on 7 September 2026 rather than copied from a feature list.
| Input | Status | What happens to it |
|---|---|---|
| Accepted | Text is extracted on the server. A scan with no text layer is read as page images instead. | |
| Word (DOCX) | Accepted | Text is extracted on the server. |
| Images — PNG, JPEG, WebP, GIF | Accepted | The picture is read directly, so a photo of a page or a whiteboard works. |
| Plain text — TXT, MD | Listed in the app, rejected in practice | The upload check turns these away. Save the notes as PDF or DOCX for now. |
One file per generation, up to 20 MB. HEIC photos straight from an iPhone camera are also refused, so export as JPEG first. The PDF row is the one we ran start to finish; the others were checked against what the app and API accept.
Two ways to generate
From a file
Pick Definitions to pull terms and their meanings out of a handout, or Vocabulary to pull lexical items and translate them into a language you choose. The deck name is suggested from the document.
From a description
Describe the deck instead of uploading anything: a topic or free-text prompt, the language for each side of the card, a CEFR level for language decks, and how many cards you want — anywhere from 5 to 100.
What the workflow actually looks like
Three screens from the test run described above. Nothing is staged, and nothing is saved to your library until you tap Create.
The complete output, not a highlight reel
All 17 cards from that PDF, in the order the app returned them. Two of them — "Role of Ca2+ in synaptic transmission" and "Axon diameter and conduction speed" — use a heading as the front rather than a question, which is exactly the sort of thing the review step is for.
Cards come back shuffled rather than in document order, so a deck never mirrors the shape of the file it came from.
| Front | Back |
|---|---|
| Synaptic cleft | The gap between the presynaptic and postsynaptic cell into which neurotransmitter is released. |
| Voltage-gated potassium channels | Channels that open during repolarisation, allowing K+ to flow out of the neuron and restore the resting potential. |
| Repolarisation | The phase where sodium channels close, voltage-gated potassium channels open, and K+ flows out of the cell. |
| Role of Ca2+ in synaptic transmission | Ca2+ enters the axon terminal, causing vesicles to fuse with the membrane and release neurotransmitter into the synaptic cleft. |
| Sodium-potassium pump | A pump that maintains the ion gradient using ATP, moving 3 Na+ out of the cell for every 2 K+ it moves in. |
| Absolute refractory period | A period during which no stimulus, regardless of strength, can trigger another action potential. |
| Neurotransmitter clearance | Neurotransmitter is cleared from the synaptic cleft by reuptake or enzymatic breakdown after binding postsynaptic receptors. |
| Threshold (action potential) | About -55 mV; depolarisation that reaches this level triggers an action potential. |
| Hyperpolarisation | A brief period where the membrane potential falls below the resting potential before returning to -70 mV. |
| Voltage-gated sodium channels | Channels that open during depolarisation, allowing Na+ to flow into the neuron and drive the rising phase of the action potential. |
| Axon diameter and conduction speed | Larger axon diameter increases the conduction speed of action potentials. |
| Resting membrane potential | About -70 mV, with the inside of the neuron negative relative to the outside. |
| Nodes of Ranvier | Gaps in the myelin sheath along an axon where the action potential signal jumps during saltatory conduction. |
| Depolarisation | The phase of the action potential where voltage-gated sodium channels open, Na+ flows into the cell, and membrane potential peaks near +30 to +40 mV. |
| Saltatory conduction | Conduction in myelinated axons where the signal jumps between nodes of Ranvier, increasing conduction speed. |
| All-or-none principle | The principle that a neuron either fires a full action potential or does not fire at all, with no partial responses. |
| Relative refractory period | A period during which a stronger than normal stimulus is required to trigger another action potential. |
The step-by-step version of this run — the original notes, which cards we kept, which we rewrote and why — is in how to turn notes into flashcards.
Review before you study
Five minutes here saves weeks of drilling a card that was subtly wrong.
- 1Read every front on its own. If it reads like a heading rather than a question or a term, rewrite it.
- 2Check any number, date or definition against the source you uploaded. The model works from your document, but it can still compress a nuance out of a sentence.
- 3Delete cards that ask the same thing twice. Long documents produce near-duplicates.
- 4Split answers longer than about two lines into separate cards.
- 5Fix the deck name. The suggested name comes from the document, not from how you file your subjects.
What it costs
A free account generates three decks per week. The count resets at the start of each calendar week, on Monday at 00:00 UTC. Premium raises the ceiling to 200 generations per week.
Studying, editing and re-studying cost nothing. A generation that fails is refunded to your allowance rather than burned.
Where your file goes
The upload travels to the Memor More API, stays in memory for the length of the request, and is not written to our database or file storage. Text is extracted from PDFs and Word files on the server, then the content is processed by Anthropic's Claude API to write the cards.
Server logs record the request and the file name, not the contents of your document. The finished deck lives on your device. More in the privacy policy.
Where it falls short
- Generation needs an account and a connection. It runs in the full native app on iPhone, iPad and Mac — not in the browser app.
- A long document is sampled, not read exhaustively. A focused chapter or lecture beats a 400-page textbook.
- The model works from your file, but it can still flatten a caveat or misread a bad scan. Cards are a draft until you have checked them.
- Plain-text uploads and HEIC photos are refused today, and the in-app counter labels the weekly allowance "this month".
- If your material is already someone else's well-made deck, browsing the public decks is faster than generating your own.
Frequently asked questions
- What files can Memor More turn into flashcards?
- PDF, Word (DOCX) and image files — PNG, JPEG, WebP and GIF — up to 20 MB, one file per generation. A PDF with a text layer is read as text; a scanned PDF is read as page images instead. Plain .txt and .md files appear in the app's list of supported formats but are currently rejected by the API, so use PDF or DOCX until that is fixed (checked 7 September 2026).
- Can it make flashcards from a PDF?
- Yes, and this is the input we tested end to end. On 7 September 2026 a one-page PDF handout of just under 200 words produced 17 question-and-answer cards in about 24 seconds on Memor More 1.3.1. Long PDFs are sampled rather than read line by line, so a focused chapter works better than a whole textbook.
- How many AI generations do I get for free?
- Three per week on a free account. The allowance resets at the start of each calendar week (Monday, 00:00 UTC). Premium raises it to 200 per week. Editing, studying and re-studying cards you already generated never counts against the limit.
- Can I edit the cards the AI writes?
- Yes, and you should. Generated cards appear in a preview inside a deck that has not been saved yet: you can regenerate the whole set, or create the deck and then edit, duplicate or delete any individual card. We verified the edit-and-save round trip on 7 September 2026.
- Does the AI generator work in the browser app?
- No. AI generation runs in the full native app on iPhone, iPad and Mac. The early browser app can create, import, edit and study decks stored in that browser, but it does not generate cards yet.
- What happens to the file I upload?
- It is sent to the Memor More API, held in memory for the length of the request and not written to our database or file storage. PDF and DOCX text is extracted on the server, then the content is processed by Anthropic's Claude API (claude-sonnet-5) to write the cards. Server logs record the request and the file name, not the document's contents. The finished deck is stored on your device.
Keep reading
- How to turn notes into flashcards → — the full worked example behind the sample deck above.
- Atomic cards and the minimum information principle — what to fix when a generated answer runs long.
- What active recall is, and why card wording matters
- Free public decks — start from someone else's deck instead.
- Memor More for students
Generate your first deck
Free to download, three AI generations a week without paying, and every card editable before you study it.
Download on the App StoreiPhone, iPad & Mac