Speaking & listening assessment
Twenty minutes. Forty spoken answers. One score on a 20–80 scale, mapped to the CEFR, with a report that tells you what it means.
How it works
Not the four-skills grid. There is no reading section and no writing section. Every answer is spoken aloud, and the whole assessment takes about twenty minutes.
01 · Listen
There is no replay and no going back. Headphones matter: without them your microphone may pick up the question audio.
02 · Speak
Not on a timer. The field on screen answers the only question you have while recording: is it hearing me.
03 · Measured
Your answers are transcribed and scored 20–80, then mapped to CEFR bands A1 to C2. Not a percentage or a pass mark. A reading.
The assessment
Short answers first, so the microphone and the nerves settle. The two open tasks come last, when you have found your voice.
Your result
A donut implies a whole being divided up. This score is a position on a 20–80 scale, so it is drawn as one: an open arc with the CEFR boundaries marked.
Provisional scale. The 20–80 transform has not yet been validated against human raters. The report says so, rather than letting the visualisation imply a precision it lacks.
Integrity
The examination environment is monitored so that a score means the same thing for every candidate. Everything it does, it does without interrupting your answer.
One photograph at the start. From then on the system only checks that you are the one still in the frame, and that you are alone.
The assessment runs in a fullscreen examination window. Leaving it is recorded; the test is the only thing on your screen.
A chime during a spoken answer would land in your recording and be scored against you. So nothing here makes a sound.
Use headphones, find a quiet room, and speak as you normally would. The instrument does the rest.