The
SproochePrett
Experience

A judgment-free practice environment with an AI proctor

Step one of the SproochePrett app: selecting a Luxembourgish topic discussion, picture description, or full oral exam simulation.

SproochePrett is an AI-powered simulator designed to mimic the oral exam format of the Sproochentest. You speak, and our native-trained Luxembourgish AI listens, transcribes your input, and provides feedback on your pronunciation and grammar. It’s a conversation partner you can practice with anytime, anywhere.

Step two of SproochePrett: practicing verbal Luxembourgish by speaking directly to a judgment-free AI exam proctor.
Practicing the Sproochentest picture description section using a timer and guided vocabulary hints to build oral fluency.

Is SproochePrett right for me?

  • SproochePrett is optimized for English-to-Luxembourgish learners who completed A2.

    It’s designed for learners who are 2-3 months out from taking the Sproochentest.

  • SproochePrett should be one part of your exam preparation, but not your sole source of prep.

    Combine it with other courses, coaching sessions, and casual in-person conversation practice.

  • Practicing with an AI proctor helps learners overcome fear of making mistakes in public.

    With sustained practice, you’ll have the confidence to speak Luxembourgish long after the exam!

For the Word Nerds

Designed for adult brains

Adults learn best through active recall and high-frequency repetition in low-stakes environments. SproochePrett bridges the "fluency gap" by forcing you to produce spoken language in real-time, moving your knowledge from passive understanding to active verbal confidence.

Not a general GPT

Generic AI models often treat Luxembourgish as a dialect of German, resulting in "Frankenstein" sentences. Our system uses a specialised architecture designed in partnership with Luxembourgis language researchers to respect the unique syntax and specific evolution of Lëtzebuergesch as a standalone language.

Lux language models

We leverage LuxASR from the University of Luxembourg to process Speech-to-Text (STT), accurately capturing the specific phonemes of Luxembourgish. This high-fidelity approach ensures that the AI understands your input even in a "low-resource" language environment where standard global tools typically fail.

Expert-vetted feedback

Our feedback loops are calibrated against the official A2 CEFR standards. You receive insights that aren't just grammatically "correct," but strategically relevant to passing the actual exam.