CareerKit

Free Voice AI Mock Interview

AI speaks the question. You answer out loud. It grades you and asks the next one.

Audio is processed by your browser's built-in speech engine — we never receive a recording, only the text transcript.

Difficulty

Type

Questions

Your browser will ask for microphone permission when you start answering. Shares the daily limit with the text Mock Interview (15 turns/day per IP).

Why practice interviews out loud?

Reading interview questions and mentally rehearsing answers is not the same skill as speaking them. Out loud, you discover the rambling intro, the filler words, the answer that takes 4 minutes when it should take 90 seconds. The Voice AI Mock Interview makes you practice the way you'll actually perform: the AI speaks each question, you answer through your microphone, and it grades the transcript before asking the next one.

Everything runs on your browser's built-in speech engine — your audio is transcribed locally by the browser and only the text reaches our servers. No recording is stored anywhere. Works in Chrome, Edge, and Safari; Firefox users can type answers or use the text-based Mock Interview.

How to run a voice mock interview

  1. 1

    Set up the role

    Same setup as the text Mock Interview: role, difficulty, behavioral / technical / mixed, and how many questions (4-8 for voice — spoken sessions run longer than typed ones).

  2. 2

    Allow microphone access

    Your browser asks for mic permission the first time you tap the record button. The audio never leaves your device — only the transcribed text is sent for grading.

  3. 3

    Listen, then speak

    The AI reads each question aloud (mute it with the voice toggle if you prefer reading). Tap the mic, answer as if a real interviewer were across the table, tap again to stop.

  4. 4

    Clean up the transcript before sending

    Speech recognition mis-hears things — company names, acronyms, jargon. Your spoken answer lands in an editable text box. Fix the errors, then send.

  5. 5

    Read the feedback like a coach's notes

    Same rigorous grading as the text version: what worked, what was missing, rubric bullets. The final summary scores you 0-100 with three strengths and three improvements.

Tips for spoken interview answers

  • Record yourself answering ONE question, listen back, and count the filler words. Most people are shocked. Awareness alone fixes half of it.
  • Aim for 90 seconds to 2 minutes per answer. Spoken answers feel twice as long to the listener as they do to you.
  • Pause before answering. Three seconds of silence reads as thoughtful; diving in immediately reads as rehearsed.
  • Structure out loud: "There are two parts to this — first… second…" — signposting is even more valuable in speech than in writing.
  • Practice your "what happened at your last job" answer until you can say it calmly in under 60 seconds. It's the question most likely to derail a spoken interview.
  • Do at least one voice session before any phone screen. Phone interviews are voice-only — the exact skill this trains.

Voice mock interview FAQs

Is my voice recorded or stored?▾

No. Speech-to-text happens inside your browser using its built-in speech engine. Only the resulting text transcript is sent to our servers for grading — same as if you had typed it. We never receive or store audio.

Which browsers does voice input work in?▾

Chrome, Edge, and Safari (desktop and mobile). Firefox doesn't implement the Web Speech API yet — Firefox users can type answers in the same interface, or use the text-based Mock Interview.

The transcription got my words wrong — does that hurt my grade?▾

No, because you edit before sending. Your spoken answer lands in an editable text box — fix mis-heard words (names, acronyms, jargon are the usual offenders) and then submit.

Does the AI evaluate my tone of voice or just the words?▾

Just the words. The browser converts speech to text before anything reaches the AI, so pacing, filler words that the engine drops, and tone aren't graded. For delivery feedback, record yourself separately and listen back.

Is the daily limit separate from the text Mock Interview?▾

No — they share the same budget (15 turns per day per IP) because they use the same underlying engine. A 6-question voice session uses 6-7 turns.

Can I switch between speaking and typing mid-interview?▾

Yes. Every answer goes through the same editable text box — speak some answers, type others, mix within a single answer. The grading is identical either way.