This page is for you if you are scoring around 60 in Schreiben on a Modellsatz, with ChatGPT or in an app, and want to know whether that is enough. Short answer: in the exam, yes; on a practice test, not safely.
Exam or practice test: what is the difference?
In the exam, 60 points is a fixed result: the module is passed. A practice score of 60 is an estimate of what real examiners would give. An estimate can be a few points off, in either direction.
What this means: 60 in the exam = pass. 60 in practice = maybe 55, maybe 65.
Where does the uncertainty come from?
Three reasons, all well documented.
Two people never score exactly alike. In Goethe Schreiben two examiners score independently and the result is their average. A study of TestDaF (German as a foreign language, Eckes 2005) found that raters differ clearly in how strict they are, even when each of them is consistent.
AI grading is often too lenient. Studies show language models grade essays more mildly than teachers (Seßler et al. 2025) and over-rate short, polished texts (Mathew et al. 2026, preprint). In our own test ChatGPT gave a B2 text with a missing Leitpunkt 82–89 points.
Our own scoring was too lenient too. In October 2026 the GoetheCoach scoring of the time passed 43 % of the exam texts that trained raters had placed below the level. We have fixed that; see how GoetheCoach scores writing.
Why 70 points is the better target
We compared 124 real exam texts (MERLIN corpus, telc tasks) with the verdicts of trained raters. The result, by score range:
| Practice score | What the raters said | What it means for you |
|---|---|---|
| 70 % and up | over 8 in 10 texts at the level | safe, with a margin |
| 55–69 % | about half at, half below the level | coin toss |
| below 55 % | over 8 in 10 texts below the level | not yet |
What this means: between 55 and 69 % the real exam could go either way. Only from 70 % does even a stricter rating still carry you over the line.
Not representative of the Goethe exam (telc tasks, CEFR levels rather than points), but a clear pattern.
What to do in the borderline zone (55–69 %)
Do not just write another text; look closely at the last one. In the borderline zone something specific is almost always missing, and it is usually quick to find:
- Is a Leitpunkt missing, or covered by half a sentence only?
- Is the text too short? At B2 the word count says "mindestens" (at least).
- Is the register right? Du or Sie, all the way through?
- Are there sentences you have to read twice? Those cost points in Wortschatz and Strukturen.
Then rewrite the text and have it scored again. Only when two or three texts in a row are above 70 are you really ready.
The fastest 10 points: from 62 to over 70
In Goethe Schreiben each criterion only has fixed bands, so one band up is worth several points at once. Here is what that looks like at B2 (official values from the rating form):
| What you improve | Band | Points |
|---|---|---|
| Teil 1: add the missing Leitpunkt | Aufgabenerfüllung B → A | +3.5 |
| Teil 1: clearer paragraphs and connectors | Kohärenz C → B | +3.5 |
| Teil 2: fewer mistakes that get in the way of understanding | Strukturen C → B | +2.5 |
| together | +9.5 |
An example: a B2 text at about 62 points, with one Leitpunkt missing in Teil 1 and unclear paragraphs, gets over 70 with these three fixes. None of them needs new vocabulary.
What this means: the fastest points are almost always in Aufgabenerfüllung, not grammar. In GoetheCoach practice texts (2026), almost every second B2 text missed at least one Leitpunkt. Exactly what a missing Leitpunkt costs is in how Goethe writing is scored.
One score is a snapshot: why you should read your level from three texts, and which criterion is usually weakest: One practice test is not enough →
Frequently asked questions
Are 60 points enough in the Goethe exam?
Yes. In the exam every module passes at 60 points. The uncertainty only applies to practice scores, which estimate what examiners would give.
What happens if I get 58 in Schreiben in the exam?
If one examiner is below and the other above the pass mark and the average is below 60, there is a third rating. If both are below 60, the module is failed. At B1 and B2 you can retake that module on its own.
Does this apply to Lesen and Hören too?
Less so. Lesen and Hören are counted with items that are right or wrong, so there are no differences between raters. A practice score of 60 there still depends a lot on the day and the Modellsatz, so some margin helps there too.
How often should I score above 70 before I register?
There is no fixed rule. A good guide: two or three texts in a row above 70, written in the official time and without help.
Is 70 on GoetheCoach the same as 70 in the exam?
No. It is our estimate, not a grade. For B1 and B2 we use the official point values and two readings, but the result can still differ from the exam by a few points. That is exactly what the margin up to 70 is for.
Sources
- Goethe-Institut: Durchführungsbestimmungen B1 and B2, § 4.1–4.3 (as of 1 September 2025)
- Goethe-Institut: B2 Modellsatz for adults, Schreiben rating form
- Eckes (2005): Rater effects in TestDaF writing and speaking, Language Assessment Quarterly 2(3)
- Seßler et al. (2025): Can AI grade your essays?, LAK '25
- MERLIN Written Learner Corpus (Wisniewski et al., Eurac Research, CC BY-SA 4.0)
- ChatGPT vs GoetheCoach: one B2 text, graded five times
How many points would your text get?
Write a task and see the points per criterion – by the same rules this article explains.
Score my text