About This MCQ / Test Paper Generator
This free tool is built for teachers, tutors, and document creators — especially in schools and colleges — who need to turn a page of notes, a textbook photo, or even just a topic name into a ready-to-use test paper without retyping everything by hand. Use "From Image (OCR)" mode to select one or more photos at once — a multi-page chapter, several textbook pages, whatever you have — and click "Extract Text From All Images" to run OCR on all of them in one go, with each image's text appended to the box separated by a clear "--- Image N ---" divider. Or use "From Topic (Auto-Research)" mode to fetch a plain-text article on any topic straight from Wikipedia. Either way, correct any mistakes in the editable box, and the rule-based generator does the rest.
You can generate three types of questions from the same source text: MCQs (it finds sentences with meaningful key terms — proper nouns, numbers, dates, or important vocabulary — blanks out the key term, and builds four shuffled answer options using other real terms from your own text wherever possible), short questions (a natural prompt built around a key term from a sentence, e.g. "What do you understand by '…'?"), and long questions (a "write a detailed note on…" style prompt built around the dominant term in a small group of related sentences). Be clear about what this is: it is not an AI writing original explanations. For short and long questions, the answer key shows your own original source sentence(s) as the "Reference answer (from source text)" — the tool surfaces and reformats your own content into exam-style prompts, it does not compose new answers for you.
Step 2 also offers class-wise / subject-wise suggested defaults: pick a Class and the tool can auto-fill sensible starting values for MCQ/short/long question counts, total marks, and time allowed for that class tier (primary, middle, matric, or intermediate). These are clearly labelled as suggestions, not an official exam-board pattern — Pakistan doesn't have one single universal format across boards, so treat these as an editable starting point. The tool never overwrites a field you've already typed into by hand; use "Reset to Suggested Pattern" if you want to reapply the suggestion anyway.
The result is a print-ready question paper with a school letterhead, a name/roll-number line for students, your instructions, and clearly separated, numbered sections — Section A (MCQs), Section B (Short Questions), Section C (Long Questions) — plus a separate answer key you can print on its own. A "Fit to page(s)" control and a live estimated-page count let you compress the paper onto one, two, or three printed pages before you print. There's no paid AI API involved — OCR, question generation and the print layout all run locally in your browser and your notes are never uploaded to a server; the only outside request is the Topic mode's fetch to Wikipedia's free public API (not a general internet search).
Frequently Asked Questions
Is this MCQ / test paper generator free?
Yes — it's completely free to use, with no signup, no watermark, and no limit on how many test papers you can generate.
Does it use a paid AI API to write the questions?
No. MCQs, short questions and long questions are all generated using rule-based algorithms that run entirely in your browser — key terms and candidate sentences are found in your own text, then turned into question stems or prompts using fixed templates. There is no external AI API call and no cost per question, and short/long "answers" shown in the key are your own original source sentences, not AI-composed explanations.
How accurate is the OCR?
OCR accuracy depends on the photo — clear, well-lit, horizontal text from printed notes or a textbook works best. Handwriting and blurry or low-resolution photos are harder to recognize correctly, which is why the extracted text appears in an editable box before you generate questions.
Can I upload more than one image at once?
Yes — select or drag & drop multiple images, remove any you don't want with the "×" on its thumbnail, then click "Extract Text From All Images" to OCR them one after another. Each image's text is appended to the box with a "--- Image N ---" divider so you can see where each one starts. If one image fails to process, the tool notes the failure in the text and continues with the rest instead of stopping the whole batch.
Can I edit the extracted text before generating questions?
Yes — this is an important step. After OCR runs, the text appears in an editable textarea so you can fix any recognition mistakes before clicking "Generate MCQs". Cleaner input text produces better questions.
What format do I get the test paper in?
The tool renders a print-ready formatted question paper with a letterhead, student name/roll number line, numbered questions split into Section A (MCQs), Section B (Short Questions) and Section C (Long Questions) as applicable, and a separate answer key section. Use "Print / Save as PDF" in your browser's print dialog to save it as a PDF or print it directly.
Are the suggested class/subject defaults an official exam pattern?
No — there's no single official pattern across every Pakistani school and board, so the suggested MCQ/short/long counts, marks, and time for each class tier are just a sensible, clearly-labelled starting point. Every field stays fully editable, and the tool won't silently overwrite anything you've already typed in yourself.
How does "From Topic (Auto-Research)" mode work — is it a full internet search?
No — it's not a general internet search and it doesn't use a paid AI. Topic mode fetches a plain-text article for the topic you type directly from Wikipedia's free, public, keyless API, right in your browser, and drops it into the same editable text box the OCR mode fills. If Wikipedia has no matching article, you'll see a clear error and can try a different spelling or switch to Image mode instead.
📖 Read the Full Guide: How to Turn Notes into a Test Paper in Minutes