German cloze tests,
generated on demand.
Fill-in-the-blank German passages with grammar, vocabulary or connector gaps. Pick a level and a topic — get a new Lückentext and answer key every time, no source text needed.
5 free generations · No credit card required
What a generated cloze test looks like
Try a real example — full set
Lückentext · Niveau B1 · 5 Lücken, 5 Kategorien
Mein Bruder hat letztes Jahr angefangen, in einem Restaurant zu arbeiten. Am Anfang war die Arbeit sehr anstrengend, ___(1)___ er noch keine Erfahrung hatte. Sein Chef hat ___(2)___ aber viel geholfen. ___(3)___ ein paar Wochen konnte er schon allein arbeiten. Heute ist er mit ___(4)___ Stelle sehr zufrieden. Er sagt, dass er später vielleicht ein ___(5)___ Restaurant eröffnen möchte. Dafür muss er aber noch viel lernen und Geld sparen.
1. Lücke (1) · Konnektor:
2. Lücke (2) · Pronomen:
3. Lücke (3) · Präposition:
4. Lücke (4) · Possessivartikel:
5. Lücke (5) · Adjektivendung:
Generated with the soruLab German Workbench — tap an answer to check it. Every worksheet comes with a printable answer key like this.
One real set: 5 gaps, 3 options each, and five different grammar categories. Every generation writes a new passage on a new topic.
Quality control
The checks this practice material passes
- Exam format. Tasks follow the Goethe and telc gap-fill (Lückentext) task types. All items are original material, not copies of exam content.
- Level fit. Grammar and vocabulary stay inside the selected CEFR band.
- One correct answer. Each item has exactly one correct option.
- Distractor quality. Wrong options come from the same word class or the same topic, so they test the skill, not luck.
- Explanations. Each set includes an answer key and an explanation for every item.
A set that fails a check does not reach you — the engine writes a new one. How we write and check questions
Why a generator beats a fixed worksheet
Most Lückentext practice comes from the same handful of textbook passages, which means learners eventually recall the missing word from memory rather than working it out from grammar or context — exactly the skill a cloze test is supposed to train. A generator that writes a new passage every time keeps the underlying skill in play instead of testing recall of a specific text.
soruLab's German Workbench lets you choose the CEFR level (A1–B2+) and the topic. It then writes a fresh passage with five gaps and a printable answer key in seconds. Each gap carries an English explanation of the grammar point it tests, so a wrong answer teaches something.
What the five gaps test
A generated set never spends all five gaps on one grammar point. It draws five different categories from the six below, so one worksheet covers the range a Sprachbausteine section actually samples.
| Category | Example gap |
|---|---|
| Article / possessive | Heute ist er mit ___ Stelle zufrieden. → seiner |
| Preposition | ___ ein paar Wochen arbeitete er allein. → Nach |
| Connector | Es war anstrengend, ___ er keine Erfahrung hatte. → weil |
| Verb form | Er sagt, dass er ein Restaurant eröffnen ___. → möchte |
| Pronoun | Sein Chef hat ___ viel geholfen. → ihm |
| Adjective ending | Er möchte ein ___ Restaurant eröffnen. → eigenes |
Every gap offers three options, and exactly one fits. The two wrong options belong to the same word class as the right one — a wrong case rather than a wrong word — because that is how the real exam builds its distractors. An option set of obviously absurd words trains nothing.
Common questions
What is a Lückentext / cloze test, exactly?
A Lückentext is a passage with certain words removed and replaced by blanks. The learner fills each blank from context, from a word bank, or by choosing the correct grammatical form. It is one of the most common exercise types in German textbooks and in exams such as the Goethe-Zertifikat and telc (Sprachbausteine).
What kinds of gaps does the generator produce?
Every set has five gaps, and each one targets a different grammar category: article or possessive, preposition, connector, verb form, pronoun, or adjective ending. You pick the CEFR level (A1 to B2+) and the topic; the generator distributes the categories so one worksheet covers the range rather than drilling a single point five times. For pure word-meaning practice, use the Wortschatz tool in the same German Workbench.
How many options does each gap have?
Three, and exactly one is correct. Both wrong options belong to the same word class as the right answer — a wrong case or a wrong verb form rather than an unrelated word — because that is how Goethe and telc build distractors. A gap whose wrong options are obviously absurd measures nothing.
Can I paste my own German text?
Not for this tool. The German Lückentext generator writes its own passage from the level and topic you choose, which is what keeps the grammar inside the CEFR band you asked for. If you want questions built from a text you supply, the English Workbench tools take a pasted passage.
Is this a JCloze alternative?
It covers the same need with less manual work. JCloze, the gap-fill module of the Hot Potatoes authoring software, still runs, but you supply the text and mark every gap by hand. Here you choose a level and a topic, and the generator writes the German passage, selects the gaps, builds the distractors and produces the answer key in one step.
What is the difference between a cloze test and a C-test?
A cloze test removes whole words and asks you to restore them from context. A C-test deletes the second half of every second word and leaves the first half as a clue. German universities often use C-tests for placement because they score quickly; exams such as telc Sprachbausteine use the multiple-choice cloze format this generator produces.
Is there an answer key?
Yes. Every generated cloze test includes a printable answer key, and each gap carries an English explanation of the grammar point it tests, so a wrong answer teaches the rule behind it.
Have questions before or after purchasing? We're here to help.
Ask us anything about plans, features, or getting started — we'll figure it out together.