Free, about 15 minutes, no sign-up. You need a phone to record yourself and any AI chat tool to score the result.
Part 1: Interview (4–5 minutes)
Start recording now. Answer each question out loud in 2–3 sentences.
Changing the set resets Parts 2 and 3 to match.
Part 2: Long turn (1 minute to prepare, up to 2 minutes to speak)
You should say:
Ready
01:00
During preparation, jot notes on paper. When the timer turns orange, speak and aim to fill the full two minutes.
Part 3: Discussion (4–5 minutes)
Keep recording. Give longer answers of 4–6 sentences: an opinion, a reason, and an example.
Stop recording after the last question.
Get your score
Turn your recording into text with your phone's dictation, or the voice-to-text feature in ChatGPT, Gemini or similar.
Copy the prompt below.
Paste it into any AI chat tool, then paste your transcript where marked.
Limits: speech-to-text hides pronunciation problems and tidies up grammar, so the AI can only estimate Pronunciation. Treat the result as a diagnosis, not an official band score.