Skip to content
Preptics
Start free

Digital SAT practice tests

Full-length adaptive mocks that behave like the real thing: two sections, two modules each, the second module routed by how you did on the first, and the same clock. You get a score range rather than a single number, a breakdown of which skills cost you points, and every explanation for every answer choice — including the ones you got right.

Questions
98
Time
134 min
Score scale
400–1600
Structure
2 sections · 2 adaptive modules each

Practice exams

What a full-length adaptive mock actually is

The Digital SAT is not a fixed test that everyone sits identically. It is a two-stage adaptive test, and the adaptation is the part most practice material gets wrong. You sit four modules in one sitting — two for Reading and Writing, two for Math — and the second module in each section is chosen by how you performed on the first.

  1. Reading and Writing, module 1

    27 questions, 32 minutes. Everyone gets the same difficulty mix.

  2. Reading and Writing, module 2

    27 questions, 32 minutes. Easier or harder, decided by how you did on module 1. This is the routing step.

  3. Math, module 1

    22 questions, 35 minutes. Built-in calculator throughout, as on the real test.

  4. Math, module 2

    22 questions, 35 minutes. Routed the same way, independently of Reading and Writing.

Reading and Writing module routingEveryone sits the same first module of 27 questions in 32 minutes. The test then routes to one of two second modules: a harder one with a ceiling of 800, or an easier one whose ceiling is below 800. Both second modules have the same question count and time limit as the first.Module 127 questions · 32 minSame for everyoneroutedonceModule 2 — harder path27 questions · 32 minCeiling: 800Module 2 — easier path27 questions · 32 minCeiling: below 800Separate conversion tables — the sameraw count scores differently
One routing decision per section, taken at the module boundary. Nothing you do inside a module changes which module you are in.

Between the two Reading and Writing modules and the two Math modules there is a ten-minute break, and the whole sitting runs about two hours and fourteen minutes. Our mocks run the same clock, the same module boundaries and the same routing rule. A practice test that is not adaptive is not measuring the same thing as the test you are going to sit — it can tell you whether you know the content, but it cannot tell you what you would score.

A Reading and Writing question in the exam runner: the passage on the left and the question with four answer choices on the right, the module name and the countdown at the top, Mark for Review above the question, and the question menu, Back and Next along the bottom.
A Reading and Writing module, on the clock the real test uses.

Why you get a range instead of a number

Every practice platform reports a score. Almost all of them report a single number, and almost all of those numbers are more confident than the evidence behind them supports. Converting raw correct answers into a 400–1600 scaled score requires the College Board's equating data. That data is not public. Nobody outside the College Board has it, including us.

So we calibrate our own conversion tables, one per second module, and then we widen the result on purpose. After one mock you get a band about 60 points wide either side of the estimate. After two it narrows to ±45, and after three or more to ±40. The band tightening is not cosmetic — it is what more evidence actually buys you.

A number like “1310” from a practice test is a guess wearing a lab coat. “1270–1350, moderate confidence” is the same guess told honestly, and it is more useful: it tells you whether the school you are aiming at is inside your range or outside it, which a false point estimate cannot.

How a projected score band narrows with more mocksA horizontal score axis from 1100 to 1500. After one mock the reported range is 1230 to 1350, described as low confidence. After two it is 1245 to 1335, moderate confidence. After three or more it is 1250 to 1330, good confidence. The centre of every range is 1290.11001200130014001500estimate 1290After 1 mock±60low confidenceAfter 2 mocks±45moderateAfter 3 or more±40good
The band stops narrowing at ±40. What remains at that point is calibration error, and watching you sit a fourth mock does not reduce it.

The full scoring method, step by step

What is free, and where it stops

Two complete adaptive mocks are free, with no card and no trial clock. Free does not mean crippled: you get the full score band, both section scores, the skill-by-skill breakdown, and every explanation for every answer choice — not just the correct one. You also get the study planner and a place on the global leaderboard.

That is how it will work at launch. Right now there is no line at all: Preptics is in beta, and every mock, lesson and drill is free for every account. At launch the pass will add volume and depth — every mock as it is released, all lessons and unlimited skill drills, and analytics across every mock — for $9.99 for three months or $24.99 for twelve. Neither will auto-renew.

Start a free mockCompare the passes

Not ready for an account? Try ten questions first — timed, marked, nothing saved. Or use the free graphing calculator and the other interactive tools.

When this is the wrong tool

Three cases where you should use something else, said plainly rather than buried:

  • You want the real, leaked question bank. Those are the College Board's copyrighted work, we do not have them, and memorised banks stop working the moment the test rotates.
  • You want a single confident number after one attempt. We will give you a range, and after one mock that range is ±60 points wide.
  • You have not booked a test date and do not intend to. The planner works backwards from a date; without one it has nothing to plan against.

If you want free foundational instruction rather than measurement, Khan Academy's official course is genuinely good and costs nothing. We are for the part after that: finding out where you actually stand, and what to do about it in the weeks you have left.

Common questions

Is the adaptive format up to date with the real Digital SAT?

Yes — the mock structure matches how the College Board actually runs the test today: two 32-minute Reading and Writing modules and two 35-minute Math modules, with the second module in each section routed by how you did on the first. If the College Board changes the format, we update the engine rather than leave you practising against a test that no longer exists.

How does the score and skill analysis work?

Every mock is scored against your actual answers, correctness and time spent, rolled up by SAT skill rather than just by section. You get a skill-by-skill breakdown of exactly where points are being lost, a score band with a stated confidence instead of a fake-precise number, and an estimate of how well you know each skill, shown as a range. Your study plan rearranges itself from those estimates after every set of drills, lesson and mock. Every calculation is explained, with its thresholds, on the How Preptics works page.

Is AI used on my data?

In two places, both written words rather than numbers. After you submit a mock, an AI model drafts the wording of your debrief from a summary our software has already calculated — score ranges, skills, pacing — which carries no name, email or account. And you can ask for a one-sentence explanation of a phrase in lessons, drills and review; that phrase is sent and nothing is kept. Neither is available in a timed test, and requests go only to providers that have committed to keep nothing, never to DeepSeek's own service. Your scores, skill estimates and plan are calculated by our own software. The How we use AI page lists everything, including limits.

How does the study plan decide what I do next?

From your skill estimates, which change as you answer. Lessons go where your estimated mastery is lowest; drills put questions due for review first, then the skills where your recent results are moving — where practice is paying off or something is slipping — with a nudge towards skills you have barely tried. After a set of drills, a lesson or a mock, the plan checks whether anything changed and rearranges itself if it did, without undoing work you have finished, and your daily mission tells you what moved.

Khan Academy and Bluebook are free. Why would I pay anything?

You should use both — they are good, and Bluebook is the real thing. What they do not do is tell you what to study today, track your proficiency skill by skill, or show you how you rank against everyone else preparing. Bluebook administers a test; Khan Academy teaches a course. Preptics is the daily routine between them. During the beta all of it is free. After launch two mocks stay free, and if the rest is not worth $9.99 to you, we would rather you used those than felt misled.

Are the mocks actually adaptive, like the real Digital SAT?

Yes. Reading and Writing is two 32-minute modules and Math is two 35-minute modules, and in each section the second module is selected by how you did in the first — the same multistage structure as the real test. Each module has its own clock held on our server, and once a module closes you cannot go back to it.

Why do you give me a score range instead of a score?

Because a range is the honest answer. Turning a raw count into a 400–1600 score requires equating against a reference population, and any new platform reporting you a precise number is guessing at a precision it does not have. We show a band and how confident we are, and the band narrows as you sit more mocks. You can read exactly how it is calculated.

What happens after I finish a mock?

A projected Reading and Writing score, a projected Math score and a total — each as a band with a confidence level. Then a skill-by-skill breakdown of where the points went, your pacing against the real module limits, and your rank this season. Then you can review every question with an explanation for each answer choice.

Is Preptics only for the SAT?

Today, yes. Preptics is a platform for standardised test preparation, and the Digital SAT is its first track: we are building it properly before starting another. PSAT/NMSQT and then the ACT are planned as the next tracks, and phone and tablet apps are planned after the web app. None of them has a date. If you need a different exam, ask for it on the feedback page, because requests change the order.

Are the questions written by AI?

Yes, and we would rather tell you than have you guess. Questions are written with Anthropic's Claude models and then put through seven checks before publication — including being solved blind by three models that never see the answer key, and Math verified by executing the arithmetic rather than trusting it. Anything ambiguous, mis-keyed, duplicated or off-specification is discarded. Then a person reviews each question and its answer key: our content auditor, an admission-test instructor who prepares students for the SAT and other standardised tests, has reviewed every question in the published practice tests. If one still looks wrong, report it — three reports take a question out of circulation automatically. One honest limit: the three blind solvers are all Claude models today, so they can share a blind spot, which is why a person checks too.

Are these real SAT questions?

No, and deliberately not. Real SAT questions and passages are the College Board's copyrighted work. Every question and every reading passage here is original: passages are written from public-domain and openly licensed source material, with the provenance recorded.

What if a question is wrong?

Report it. Every question has a report button, and once a question crosses a report threshold it is automatically pulled from circulation while we re-verify it. If we remove a question, affected scores are recalculated so removing it can never lower your score.

More in the full FAQ and the help centre, or read how our questions are written and verified.

SAT® is a trademark registered by the College Board, which is not affiliated with, and does not endorse, this site. All practice questions and reading passages are original works created by Preptics and are not actual SAT® questions.