DOCS · FEATURES
Speaker Mode
Listen to your interviewer through the speaker, get voice-to-text in real time with no latency, and answer in one tight sentence. Built for fast theory and behavioral Q&A.
How to use it
- Press Ctrl + Shift + A to start listening.
- As your interviewer speaks (over Zoom / Teams / Meet through your speaker output), the live transcript appears above the chat input — word-by-word, real time, no perceived latency.
- When they finish the question, press Ctrl + Shift + A again. The transcript is sent to the AI, which fires back a one-sentence answer you can read aloud.
When this is useful
- Theory questions: "What's the difference between TCP and UDP?", "Explain the CAP theorem", "What's a Bloom filter?"
- Behavioral warm-ups: short, definition-style verbal questions where a multi-paragraph answer would be wrong.
- Definition checks: when the interviewer uses a term you want to be sure you're aligned on.
Customize the answer style
Settings → General → Speaker instruction. Default is "answer in exactly one tight sentence" — change it to bullet-list, longer answers, or a specific persona if you prefer.
Tips
- If the interviewer's mic is quiet, transcription accuracy drops. Bump your system-audio volume in Windows sound settings.
- Press Esc after toggling off to cancel the AI response if you didn't mean to send.
- Free trial accounts get 5 minutes total to try Speaker Mode. Subscribe for unlimited.