A real-time interview assistant is a desktop app that captures the interviewer's voice from your meeting audio, shows live captions, and puts a structured answer on your own screen in about 2 seconds after the question ends. This page collects the short questions people ask about that kind of tool and answers each in a sentence or two, using CoPilot Interview as the concrete example. For the long version, start with our real-time interview copilot page.
Setup
What do I need to run one?
A Windows 10/11 or macOS 12+ computer, an internet connection, and a meeting app such as Zoom, Microsoft Teams or Google Meet. CoPilot Interview downloads as a .zip from copilotinterview.com or installs from the Microsoft Store.
Does it need my microphone?
Yes. In a live session the app captures two sources and mixes them: the computer's audio, meaning what the meeting app plays, which carries the interviewer, and your microphone, which carries your own answers so the 📊 Summary report can score what you said. A Mute mic toggle keeps only the interviewer's audio if you prefer. The mock interview and the Android companion app use the microphone only.
Can I use headphones?
Yes, and they are usually better because they remove echo. The interviewer comes through the computer's audio while your microphone captures you, so headphones change nothing about what the app hears from the interviewer and keep their voice out of your mic.
Is there a phone app?
There is an Android companion app in beta. It listens through the phone's microphone and cannot capture call audio or screens. There is no iPhone app yet.
Do I have to create an account?
No. The Free plan needs no signup and no credit card.
How do I set the language?
Pick your interview language in the app; captions and answers are available in dozens of languages including English, Spanish, French, German, Portuguese, Hindi, Chinese, Japanese, Korean, Arabic and Russian. Since version 1.3.16, if the interviewer speaks a different language from the one set, the app offers a one-click switch within the first few sentences.
During the call
How fast does the answer appear?
About 2 seconds after the interviewer stops talking. Key points come first, then detail, so you can start speaking from the first line.
Do I have to click anything?
No. Auto-answer can trigger after the interviewer pauses, and a new question is detected automatically after a 20-second pause in speech. Global hotkeys also work while your cursor stays in the meeting app or a shared editor.
| Hotkey (Windows; ⌘ replaces Ctrl on Mac) | What it does |
|---|---|
| Ctrl+Shift+L | Listen or pause |
| Ctrl+Shift+A | Answer the current question |
| Ctrl+Shift+R | Redo the answer |
| Ctrl+Shift+Q | Start a new question |
| Ctrl+Shift+S | Read screen |
| Ctrl+Shift+C | Coding mode on or off |
| Ctrl+Shift+E | Talk answer: words instead of code, for one question |
| Ctrl+Shift+H | Show or hide the app window |
| Ctrl+Shift+/ | Scroll the answer |
What if the answer is too long?
Press the shorter-answer button, or ask for bullet points. The longer-answer button does the opposite when the interviewer wants depth.
What if the answer is wrong for me?
Press Redo (Ctrl+Shift+R) or type your own question. You are the one answering; treat the text as a structure and fill it with what you actually know.
Can it write code?
Yes. Coding mode returns complete code in your chosen language, with 50+ languages in the picker and an example run. A 📝 Note option adds a 2-4 line approach and complexity comment at the top; 🚫 No comments gives clean code.
What about a conceptual question in a coding round?
Press 🗣 Talk answer (Ctrl+Shift+E). The next question is answered in words while Coding stays on.
Can it see the shared editor or a slide?
Read screen (Ctrl+Shift+S) and Pin screen capture a shared coding editor or a slide and answer what is shown. 🐞 Debug mode finds bugs in code captured from the screen.
What if I accidentally click into the window?
Turn on click-safe (Ctrl+Shift+F). The window then never takes focus from the app you are working in.
Cost and plans
Is there a free version?
Yes. The Free plan is $0 forever: 15 minutes a day in sessions of up to 15 minutes, free AI models, and every mode including the mock interview and the résumé builder.
What does a paid plan add?
Mostly live minutes and models. Standard ($14.99 a month, regular $19.99) gives sessions up to 60 minutes and up to 100 minutes a day with no waiting between sessions, plus Dual AI and Focus Mode. Pro ($29.99 a month) has 70-minute sessions, unlimited sessions, and the ChatGPT and Gemini models. Pro+ ($49.99 a month) adds Claude.
Can I pay for one interview without a subscription?
Yes. Hour packs never expire: 1 hour for $5.99, 5 hours for $19.99, 15 hours for $34.99.
Is there an annual option?
Standard is $99.99 a year, Pro $199.99, Pro+ $249.99. Details are on the pricing page.
Is there a refund?
A 14-day money-back applies to the first payment of a new subscription if under 30 minutes of live session time was used.
Privacy
Is my interview recorded?
No. Audio is streamed in short encrypted segments through CoPilot Interview's proxy to a speech-to-text model, and no audio, transcripts or answers are stored on CoPilot Interview's servers.
Where does the transcript go?
Session files (transcript, answers, report) are saved in your own Documents folder, on your computer.
Does the app send anything else?
Optional anonymous question statistics (titles, categories, scores — never answers, transcripts, résumés or contact details) can be turned off in Options. The security page has the full description.
Which AI model sees my question?
The one you picked. Free and Standard use fast free models (Mercury 2.5 by Inception is the default; GPT-OSS and Qwen are available). Pro adds ChatGPT (GPT-6 Luna) and Gemini (2.5 Flash-Lite); Pro+ adds Claude (Haiku 5.5) and Gemini 3 Flash. You can switch mid-session.
Limits
Does it work offline?
No. Speech-to-text and the AI models run in the cloud.
Does it work in an in-person interview?
Not live. For interviews in a room, the useful part is the AI mock interview, which builds questions from your job description and résumé and scores your spoken answers.
Why did captions stop after a quiet stretch?
An idle pause stops recording after 10 minutes with no voice. Press Ctrl+Shift+L to resume.
Does it know my experience?
Only what you paste in. With your résumé and the job description loaded, answers reference your real projects; without them, answers are generic and you have to supply the specifics yourself.
Is it allowed?
That depends on the employer. Check the invitation and ask the recruiter if it is unclear. Some employers say AI tools are fine, some ask you not to use them, and many say nothing. If you are weighing tools, you can compare AI interview assistants on exactly these questions before you pick one; the broader AI interview assistant page covers the practice and reporting modes that sit behind the live window.
Follow the rules of the employer or assessment you're in — many people use it only to practice, which works on the free plan.
Answer the rest of your questions by trying it
Open the app beside a test call, play a video with speech, and watch the captions and answer appear. Free plan, no signup, 15 minutes a day.
Download freeFAQ
How does a real-time interview assistant hear the interviewer?
It captures the computer's audio, meaning whatever the meeting app plays, so the interviewer's voice does not depend on the room and headphones work normally. Your microphone is mixed in alongside it to capture your own answers, with a Mute mic toggle if you want the interviewer only. If the call is on a phone instead of the computer, there is no computer audio for the desktop app to capture.
How long does it take to get an answer during the call?
About 2 seconds after the interviewer stops speaking. The answer is structured with key points first and detail below, so you can begin talking from the first line while the rest loads.
Can I use a real-time interview assistant for free?
Yes. CoPilot Interview's Free plan is $0 with no signup and no credit card, giving 15 minutes a day in sessions of up to 15 minutes. Every mode is included, with the free AI models.
Does the app keep my interview audio or transcript?
No audio, transcripts or answers are stored on CoPilot Interview's servers. Audio is streamed in short encrypted segments for transcription only. Your transcript, answers and report are saved in your own Documents folder.
What happens if the interviewer switches languages?
Since version 1.3.16, if the interviewer speaks a different language from the one you set, the app offers a one-click switch within the first few sentences. Captions and answers are available in dozens of languages.
Does it work for in-person interviews?
Not live, because there is no screen to read from across a table. The practice side, the AI mock interview, is the part that helps for in-person rounds. The Android companion app is designed for calls, not rooms.