B1 speaking usually has three distinct parts: planning something together with a partner or examiner (a party, a trip), a short prepared presentation on an everyday topic, and reacting appropriately to a request or complaint in a short roleplay. The single biggest score-killer is not grammar mistakes — examiners expect some at B1 — it is silence: stopping mid-sentence because you do not know a word, instead of describing around it. Asagno's speaking practice uses the same AI roleplay engine built for real-life scenarios (apartment viewing, doctor visit, job interview), set to B1 level, so you rehearse actually talking, not just reading a script — with a "what can I say?" hint when you get stuck and a full report afterwards covering your objectives, mistakes and suggested phrases.
What you'll learn
- The three-part B1 speaking format: joint planning, a short presentation, reacting in a roleplay
- How to describe around a word you do not know instead of going silent
- Phrases for agreeing, disagreeing politely, and making a suggestion
- AI roleplay practice at B1 level, by text or voice, with a "what can I say?" hint
- A full end-of-session report: objectives met, mistakes with corrections, better phrase suggestions
Rescuing a sentence mid-way
Forgot the word for "landlord"? Say "die Person, die die Wohnung vermietet" (the person who rents out the apartment) instead of stopping. Examiners consistently reward this kind of circumlocution over silence.
Frequently asked questions
Is the speaking test with a real person or a computer?
The real exam is with a human examiner (often paired with another candidate). Asagno's AI roleplay is practice for that — the format and objectives match, and it never claims to be the real exam.
Can I practise by voice, not just text?
Yes — every roleplay scenario supports both text and voice.