WeSolve+ is best for a student with their own notes or PDFs who wants practice out of them rather than a chat about them. The tool above starts you off with text rules and no account. Everything below is the method itself, with the measurements it rests on, and it works whether or not you use our app.
WeSolve+ reads the whole document and writes the questions for you
Upload your PDF, photograph your notebook, or point the camera. WeSolve+ writes questions from that material, explains why each answer is right, reads the chapter back to you as a podcast, and remembers every item you missed until you own it.
The tool below is a small browser-only tool and it is not WeSolve+: paste a few lines and text rules turn them into cards on the spot. The real app, the one that uses AI, is behind the link above.
Turn notes into practice
This is a browser-only tool, and that is all it isIt splits the text you paste by rule, and nothing else. WeSolve+ is a different thing entirely: it reads your whole PDF with AI, writes the reasoning behind every question, speaks the chapter back to you, and remembers what you missed so it can return it. Try the real app now, free!
This tool uses text rules, not AI. It looks for lines written as "term: definition" or "term = definition" and makes a two sided item; for any other line it removes the longest word and asks you to supply it. Your text stays in the browser and is never sent anywhere. The copy button produces tab separated rows, the format Anki and Quizlet both accept.
The one rule: study your material, not the model's
A model asked to explain photosynthesis will produce a fluent answer that may not match your syllabus, your lecturer's emphasis or the exam you are sitting. A model asked to build questions out of your lecture notes cannot drift, because the source is fixed. That is the whole distinction, and it is why every output in WeSolve+ is tied to an uploaded file rather than to a prompt. It is also why the tool at the top of this page has no knowledge of biology: it cannot invent a fact because it was never given one.
Testing yourself beats reading it again
Roediger and Karpicke put numbers on this in 2006: after a week, students who had taken a practice test on a passage remembered clearly more than students who had reread it, even though rereading felt more productive while they did it. Karpicke and Roediger found the same direction in 2008 against repeated study. The catch is the feeling. Retrieval is uncomfortable, rereading is smooth, and people reliably choose the smooth one. Any AI workflow that ends with a summary you read is the smooth one wearing a new coat.
Spacing the reviews, and how far apart
Cepeda and colleagues reviewed 254 studies in 2006 and found spacing the same total study time across days consistently beat massing it. In 2008 they addressed the harder question of how far apart, and the answer scales with how long you need to keep the material: a test next week wants shorter gaps than a test in three months. You do not need to calculate this yourself if your tool schedules it, which is what spaced repetition does, but you do need to know that one long session before the exam is the pattern the evidence argues against.
A workflow that fits a real week
Upload the week's material once and let it produce a quiz, a card deck and a short audio recap from the same file. Take the quiz the same day, while the lecture is still fresh; the wrong answers collect in their own list. Two days later, review only the wrong ones plus whatever the scheduler surfaces. Listen to the recap on a walk when you have no hands free for a screen. Before the exam, run the quiz again cold rather than rereading the notes. Nothing in that loop requires a particular app, and the longer guide writes it out step by step.
What AI is bad at here, plainly
It is bad at deciding what matters in a subject it cannot see the syllabus for. It is bad at figures: a question that refers to a diagram you were not shown is useless, which is why generated questions are built to be answerable from text alone. It is bad at knowing you personally, so a plan generated on day one is a guess until it has seen you get things wrong. And it is bad at telling you when it is wrong, which is the argument for keeping every output tied to a source page you can check.
Starting without paying anything
The tool above is free, has no usage cap and needs no account, because it never contacts a server. The app is free to start in your browser at wesolveapp.com and as an App Store app on iPhone and iPad: 2 quizzes per day, 2 PDF uploads per day, 2 audio recaps per week and 2 duels per day on the free plan. There is no separate Android app; on an Android phone you use the browser. Premium and Ultra raise the daily limits and are managed through the App Store. Details are on the pricing page.
Sources used on this page
- Roediger and Karpicke, Test-Enhanced Learning (2006)
- Cepeda et al., distributed practice meta analysis (2006)
- Dunlosky et al., improving students' learning (2013)
- Hermann Ebbinghaus, the forgetting curve (1885)
- Anki manual, importing text files
- Karpicke and Blunt, retrieval practice and concept mapping (2011)
- Bjork and Bjork, desirable difficulties
- WeSolve+ on the App Store
- MDN, the Clipboard API
- MDN, the details element
- Karpicke and Roediger, the critical importance of retrieval for learning (2008)
- Cepeda et al., spacing effects and the optimal retention interval (2008)
- Kornell and Bjork, is spacing the enemy of induction (2008)
- What Works Clearinghouse, organizing instruction and study
- The Learning Scientists, retrieval practice
- Retrieval Practice, the research library
- Cloze test
- Active recall
- Flashcard
- Spaced repetition
- Testing effect
- Leitner system
- SuperMemo, the SM-2 algorithm
- Anki manual, studying and scheduling
- Anki manual, deck options
- MDN, the textarea element
- MDN, the output element
- W3C, the disclosure pattern
- W3C, Web Content Accessibility Guidelines 2.2
| Habit | What the evidence shows | Where it fits here |
|---|---|---|
| Rereading notes | Feels productive, retains less after a week | Replace with a quiz on the same notes |
| Testing yourself | Retains clearly more at one week | The quiz and card modes above |
| One long session | Beaten by the same minutes spread out | Space the reviews across days |
| Chatting with a model | Answer may not match your syllabus | Generate from your own file instead |
| Highlighting | Rated among the least effective techniques | Turn the highlighted line into a card |
| Listening while walking | Useful for a second pass, not a first | The audio recap |
Is studying with AI actually effective?
It depends entirely on what you have it do. Using it to generate practice from your own material and then testing yourself lines up with what the research supports. Using it to produce summaries you then read is closer to rereading, which measured worse than testing after a week.
Will AI do my homework for me?
That is not what this is for, and a generated answer you did not work through will not help you in an exam room. The workflow here is the opposite: it turns your material into questions you have to answer yourself.
Do I need to pay to start?
No. The tool on this page is free with no account. The app is free to start in your browser and on iOS, with daily limits on the free plan.
Can I use it on an Android phone?
Yes, in the browser at wesolveapp.com. There is no separate Android app. iPhone and iPad additionally have an App Store app.
How long before an exam should I start?
Earlier than feels necessary, because the benefit comes from spreading the reviews rather than from total hours. The spacing research is consistent on the direction even where the exact intervals depend on how long you need to retain the material.
Does the AI make things up?
The risk is real for any model, which is why every output here is generated from a file you uploaded and stays tied to it. A question you can trace back to a page is a question you can check.
Last updated: 2026-08-15