How we grade · In full
How LëtzPass grades your speaking practice
The exact criteria and weighting the Sproochentest examiners use are not published in any form we can verify. So anyone who claims to score your speaking exactly like the real examiners is guessing. We take the honest path instead. Here is our own practice-scoring model, in full, with a worked example, and a plain account of what it is and what it is not.
This is the model that runs inside the app, grounded in our own code. It is not the official rubric, and it is not endorsed by the INLL. For what the real exam itself assesses, read how the official exam is scored.
The seven qualities we listen for
When you finish a spoken answer, our scorer weighs these seven qualities of everyday communication. Notice what is missing: flawless grammar, rare vocabulary, clever sentences. A candidate who speaks simply and steadily reads better here than one who reaches for complexity and stalls.
Ability to make yourself understood
Whether a listener can follow your meaning, even when the language stays simple.
- Weaker
- Broken structure that is hard to follow.
- Stronger
- Short, correct sentences a listener understands the first time.
Range of vocabulary
Whether you have enough everyday words for the topic, without reaching for rare ones.
- Weaker
- Leaning on generic fillers.
- Stronger
- Naming specific, concrete things.
Mastery of basic grammatical structures
Whether your simple sentences are correct, not whether they are complex.
- Weaker
- Complex clauses that break down halfway.
- Stronger
- A correct present-tense subject and verb, kept simple.
Clarity and fluency
Whether you speak steadily, without long hesitation.
- Weaker
- Long pauses and heavy fillers that stall the answer.
- Stronger
- A steady pace, with the normal amount of A2 hesitation.
Responsiveness to the task
Whether you answer the question that was asked, or describe the picture in front of you.
- Weaker
- Off topic, which stays Not Yet even when the grammar is good.
- Stronger
- A direct answer to the actual question.
Coherence
Whether your sentences hold together and mean something, without contradicting themselves.
- Weaker
- A statement that contradicts itself, such as loving a place because it is boring.
- Stronger
- Ideas joined by a linker.
Interactivity
Whether you respond like a real back and forth conversation.
- Weaker
- A one-way delivery with no thread to pick up.
- Stronger
- Taking the prompt and carrying it into a natural reply.
In a finished report, these qualities roll up into one score from 0 to 100, a band, and a short per-answer breakdown. Here is what that looks like.
A report, annotated
This is a static recreation of a real feedback card, built from a neutral example answer. The four markers show what each part of the report tells you.
“Jo, gär. Dat ass praktesch a mécht Spaass.”
A direct, on-topic answer with a clear reason. One more concrete detail would move it from Pass toward Confident Pass.
What went well
You answered directly and gave a reason, which is exactly the beginner strategy the exam rewards.
Next step
Add one more short detail, for example who you cook with, to move from Pass toward Confident Pass.
“Jo, gär. Dat ass praktesch a mécht Spaass. Ech kachen dacks mat menger Fra.”
What each part tells you
Score and band
One number from 0 to 100, placed in a band. The band is the headline you read first: Not Yet, Pass, or Confident Pass. The pass line sits at 50.
Coaching
Feedback in the same shape every time: one thing you did well, and one concrete next step, both in plain English.
Per-answer breakdown
The same answer scored on relevance, language quality, and depth, so you can see which part to work on. The amber bar is the weakest one to target next.
Model answer
A native-speaker version that takes your answer one step further, to compare against.
What this scoring is not
Being clear about the limits is the point of the page. Our model is a practice tool, and it stays inside these lines.
Not the official rubric
The examiners' own criteria and weighting are not public. We do not reproduce them or claim to match them.
Not a prediction
A practice score does not forecast your real result. We have no study behind such a claim, so we do not make one.
Not a certificate
Nothing here is an official assessment. Only the INLL sets and marks the real exam.
The real exam, for reference
The practice mock is our own simplification, built to be quick and repeatable: it is picture-first, about 25 minutes, and speaking only. The real Sproochentest is different. Speaking lasts 10 minutes in two parts, the interview first and then a picture description, assessed at A2. Listening is a separate 35-minute section at B1. To pass, you need at least 50 percent on the speaking test, or an average of at least 50 percent across speaking and listening.
What you see in the free mock
The free mock shows a partial version of the report above. You get your score, your band, and one strength and one next step, marker 1 and marker 2. The per-answer breakdown at marker 3, the model answers at marker 4, and the full transcript detail are part of a plan.
Questions about our scoring
- Is this the official Sproochentest scoring rubric?
- No. The official examiner criteria and their weighting are not published in any form we can verify, so we do not reproduce them. What you see here is our own practice-scoring model, built into LëtzPass and presented as ours.
- Does my practice score predict my real exam result?
- No. We have no study linking our score to the official outcome, so we make no prediction. A practice score tells you how a clear, on-topic answer is developing, and nothing more.
- Why is the score the same each time I give the same answer?
- Because it is the same model on every run. Scoring is consistent from one practice answer to the next, which is what makes progress readable over a few weeks.
- What can I see for free?
- One full mock exam, with your score, your band, and one strength and one next step. The per-answer breakdown, the model answers, and the transcript detail are part of a plan.