The product

One session a day, and it is already chosen.

Reading, listening, writing and speaking — the writing and speaking marked against the published criteria, with your own sentences quoted and rewritten. The route decides what today is, weighted by whatever you are currently worst at, and paced against your real exam date.

Open today’s session Find your level first

  • One buttonnever a menu of hundreds of items
  • It endsabout ten minutes, and finishing is the point
  • It adaptsweighted by subskill, not by topic

Always on screen

3 numbers days to your exam, your streak, and where your readiness stands today.

The four parts, and where each one is practised

Everything below is on this page. The daily session mixes the parts and decides the balance; the full sitting at the bottom runs all four under the clock, the way the examination runs them.

Lezen — reading

Texts written for this service at B1 and B2, with questions about what is meant rather than what is written. Marked from your answers the moment you give them, with the reason the right answer is right.

Practise it in today’s session

Luisteren — listening

Recorded fragments with the questions the real paper asks. In the daily session you can replay a clip while you are learning. In the full sitting you cannot: it starts itself, cannot be paused and plays once, because that is what the examination does and it is the thing candidates most often get wrong.

Practise it in today’s session

Schrijven — writing

Real tasks with the situation, the addressee and the length the criteria expect. You get a verdict per published criterion, the sentence you wrote quoted back for every mistake that cost you a mark, why it is wrong in Dutch, and the sentence you should have written.

Sit the writing part in a full sitting

Spreken — speaking

Preparation time, then a recording that starts by itself and stops by itself, with no going back to a previous task. Your own recording comes back with markers at the moments things went wrong, and tempo is measured rather than judged.

Sit the speaking part in a full sitting

Correction that teaches, not correction that scores

A band on its own is a grade. For every mistake that cost you a mark you get the sentence you wrote, what is wrong with it, why it is wrong in Dutch, and the sentence you should have written. Adequacy is a gate in this examination — an answer that does not do what the task asked scores nothing for content and caps everything else — and the marking says so in those words when it happens.

Marked work is stored with the version that produced it

Every mark carries the version of the criteria, the version of the instructions the assessor was given, and the model that did it. That is what makes a re-mark explainable a year later and an old mark re-runnable, instead of a number with a date on it.

The full sitting is a milestone, not the daily diet

Four parts, each timed by our server rather than by your browser, and closing the tab does not hand a second back. It is a measurement, and a measurement run every week eats the practice time that would have produced the improvement it is measuring — so the route offers one when enough evidence has built up, and again as your date comes close, and never in the last days before you sit the real thing. What it costs in markings is on the screen before it starts, never discovered halfway through. Our own paper, in our own words, with nothing official reproduced.

How the daily route works

One session a day, about 10 minutes, and it ends. There is no catalogue to pick from: the route decides what today is from what you got wrong, when you got it wrong, and which of the 10 subskills you are currently weakest at.

Some days the route hands you a writing or a speaking task instead of another question. Those are marked against the published assessment criteria, band by band, with the evidence quoted from your own sentences and a corrected version of every sentence that cost you a mark — and a spoken answer comes back with markers on your own recording at the moments things went wrong. 20 marked submissions are included.

Anything you answer wrongly comes back a few days later on its own, and again later if you miss it again. The route keeps bending toward your weakest subskill until it stops being the weakest — not toward a topic, toward the thing the questions actually measure.

Enter your real exam date and the route paces itself against it, and tells you whether the pace you are actually keeping arrives in time.

A full mock sitting

Four parts, each on its own clock, in the order they are always given here: reading, listening, writing, speaking. It behaves the way the examination is documented to behave — the recordings play once and start themselves, the speaking part never goes back, and the clock keeps running whether or not this page is open.

Independent practice. Not affiliated with any examining body. This is our own paper, written against the published description of the examination — no official material is reproduced here, and no result from it is an official result.

  • Programma I (B1) — 66 questions, 8 marked pieces, four parts, about 5 hours
  • Programma II (B2) — 28 questions, 6 marked pieces, four parts, about 5 hours

What is enforced, and why it is worth knowing before the day

  • Each part is timed by our server, not by your browser. Closing the tab, losing your connection or coming back on another device does not give you a single second back.
  • A listening recording plays once, starts itself, and cannot be paused. You may go back to an earlier question, but you will not hear the fragment again.
  • The speaking part does not come back and does not skip ahead: one task at a time, and once you have answered it, it is gone.
  • You will not be told whether an answer was right until the whole sitting is finished. A running score would change what the rest of the paper measures.

Where this paper is shorter than the real one

  • Luisteren: 30 where the real paper has 40. Our own bank holds fewer recordings than a full paper needs, and a sitting that used most of them would leave a second sitting nothing new to ask.
  • Schrijven: 4 where the real paper has 12. Every written and spoken answer is marked against the published criteria and each marking costs money, so the sitting takes a representative spread of the published task types rather than all of them.
  • Spreken: 4 where the real paper has 16. Every written and spoken answer is marked against the published criteria and each marking costs money, so the sitting takes a representative spread of the published task types rather than all of them.

Stuck, or something not working as described? Write to a person and we answer within two working days.

Independent practice. Not affiliated with any examining body, and no official examination material is reproduced here.