New in PaperScorer

Describe the test. Get the questions.

Write one line about what you want to assess and PaperScorer drafts the multiple choice questions — distractors and answer key included, aligned to your standards if you want them. Up to 50 at a time, for as low as six cents a question.

Available on every plan. Tokens are a one-off purchase — no subscription.

One line in, a unit test out

There is no prompt engineering to learn. You describe the assessment the way you would describe it to a colleague.

What you type

5th grade water cycle, focus on evaporation and condensation

What comes back

  • Multiple choice questions written to that topic and grade level
  • Plausible distractors, not obviously wrong filler
  • A verified answer key, with the options shuffled
  • A link to a standard on every question, if you chose standards
  • Optional topic tags for grouping and filtering

It does not have to be a school subject

The prompt is plain text, so it will write questions about anything you can describe. Teachers use it for coursework. Trainers, coaches, tutors, and quiz writers use it for everything else.

Culinary

Knife cuts and kitchen safety for new line cooks

Sports

NFL penalties and how officials enforce them

Film

1970s American cinema, directors and their films

Workplace training

Forklift safety refresher for warehouse staff

Music

Reading rhythm notation, beginner level

Trivia night

World capitals and national flags, hard mode

How it works

Three steps, and the third one is the important one.

01

Describe what you want

Write it the way you'd say it out loud — "5th grade water cycle, focus on evaporation and condensation." Name the subject, the grade, and anything specific you need covered.

02

Set the shape of the test

Choose how many questions and how many answer options each one gets. Standards are optional — add them if you want the questions written against them, or leave them off entirely.

03

Review and edit

Questions drop straight into your assessment, linked to the standards you picked if you picked any. Every one is editable. Nothing prints until you have read it.

What it actually does

The specifics, so you know what you are buying before you buy a token.

Standards are optional, not assumed

Add the standards you are teaching and every question comes back linked to the one it addresses — either a set count per standard or one total spread across them — so mastery reporting works from the first scan. Skip standards and you simply get an untagged set.

Up to 50 questions per run

Generate a full unit test in one pass. Answer options are configurable from 2 to 10 per question, with 5 as the default.

Distractors written on purpose

Wrong answers are the hard part of a multiple choice item. The generator writes plausible distractors, shuffles the options, and checks the answer key before handing the set back.

Optional auto-tagging, free

Turn on tagging and each question gets one to three short topic tags for grouping and filtering. It is off by default and uses no extra tokens.

One token, one question

Buy tokens once and spend them whenever. The larger the package, the less each question costs — as low as six cents.

25

questions

$3.00

$0.12 per question

50

questions

$5.00

$0.10 per question

100

questions

$9.00

$0.09 per question

250

questions

$20.00

$0.08 per question

500

questions

$35.00

$0.07 per question

Best value

1,000

questions

$60.00

$0.06 per question

Available on every plan, including Free. There is no trial allotment — you buy the package that fits and the balance is yours until you spend it.

Where it fits

AI writes the questions. Paper keeps the answers honest.

We argue elsewhere that generative AI has made digital take-home work impossible to trust, and that the fix is testing on paper, in person, under supervision. That argument is about who did the work — a student handing in something a model wrote.

Who drafted the question is a different matter entirely. A well-written item is well-written regardless of what typed the first version, and a supervised paper test is trustworthy regardless of how its questions were authored. AI helping you build the test does not weaken the test.

Read the AI-resistant assessment argument

What it does not do

  • Generate anything other than multiple choice — other question types are still built by hand
  • Work on assessments you only hold an answer key for; the assessment has to be managed in PaperScorer
  • Publish anything on its own — every question is a draft until you review it
  • Know your classroom — questions that reference Tuesday's discussion still have to come from you

The last one matters most. Course-specific prompts are what make an assessment hard to shortcut — see the True Test of Knowledge approach.

Common questions

How much does it cost?

One token generates one question. Tokens are a one-off purchase in six sizes: 25 for $3.00, 50 for $5.00, 100 for $9.00, 250 for $20.00, 500 for $35.00, and 1,000 for $60.00. The largest package brings it as low as $0.06 per question — the best rate available. Tokens do not expire on a subscription cycle — you buy them and they sit in your balance until you use them.

Is it included in my plan?

AI generation is available to every PaperScorer user on every plan, including Free. It is priced separately with tokens rather than bundled, so you only pay for what you actually generate. There is no free token allotment — new accounts start at a zero balance and buy the package that fits.

What question types can it generate?

Multiple choice, with 2 to 10 answer options per question. Other PaperScorer question types — true/false, numeric, matching, ordering, and written response — are still built by hand, and you can mix them into the same assessment alongside generated questions.

Does it only work for school subjects?

No. The prompt is free text, so it will generate questions on any topic you can describe — cooking technique, football rules, film history, warehouse safety, product training, or trivia. Corporate trainers, tutors, coaches, and quiz writers use it exactly the way teachers do. Because standards are optional, a non-academic set simply comes back without them.

Do I have to generate against standards?

No. Standards are optional. If you add them, every question comes back linked to the standard it addresses, and you can either set a number of questions per standard or give one total and let it spread across the standards you chose. If you skip standards, you get the questions with no standard attached — you can still tag them yourself afterward, the same as any hand-written question.

Can I edit the questions it writes?

Yes, and you should. Generated questions land in your assessment as ordinary editable items — change the wording, fix a distractor, retag, reorder, or delete any of them. Treat the output as a first draft written quickly, not a finished test.

Does it work on assessments I imported or built elsewhere?

No. AI generation is only available for assessments managed in PaperScorer. If you are scanning against an existing paper test where PaperScorer only holds the answer key, there is no question content for the generator to add to.

Which AI model does it use?

Questions are generated by a commercial large language model accessed over an API. Your assessment content is not used to train it. If your district needs specifics for an AI review or procurement questionnaire, get in touch and we will answer in writing.

Isn't it contradictory to sell AI question writing when you argue AI undermines assessment?

They are different problems. The integrity problem is a student using AI to produce work you then grade as theirs — which is why we argue for paper, in person, under supervision. Who drafted the question is not what makes a result trustworthy; how it was administered is. AI helping you write a test that students sit on paper does not weaken that test at all.

Build your next test in a sentence

Create a free account, buy a starter pack of 25 tokens for $3, and see what comes back.