Frequently asked questions

Dertos AI is a real-time AI copilot for interviews and meetings. A small overlay transcribes the call, detects questions and streams suggested answers based on your resume, notes and the job description. It also solves coding questions from screenshots, runs mock interviews and writes notes after each call.

It listens to the interviewer's audio and starts an answer when a question ends. You can turn auto-answer off and press ⌘↩ (Answer now) whenever you like, or type a question with ⌘⇧C.

Yes. Press the screenshot hotkey to capture the problem (up to 5 screenshots). Dertos AI reads them together with the conversation and returns the question, the code and key points with complexity. It can also fix or optimise code that's already on screen.

Yes. Mock interview mode asks role-specific questions, listens to your answer and scores it, with strengths, improvements and a sample answer.

78 languages for transcription, answers and notes. Answers can follow the language of the question.

Yes: one free 10-minute session with no card needed, so you can try the full experience.

Credits are for occasional use. Half a credit covers one 30-minute call session or mock interview, so one credit is an hour. Bigger packs include free bonus credits: 8 credits for the price of 6, 15 for the price of 9.

Every feature and full stealth mode, with call sessions and mock interviews up to 7 hours a week (₹2,490), 30 hours a month (₹4,735) or 120 hours a year (₹18,950, about ₹1,579/month). Each is a one-time payment for that period: nothing renews automatically, and you buy the next one when it ends.

If you use your own Claude, OpenAI or Deepgram key, that provider bills you for usage. A free Google Gemini key, or local Ollama and Whisper, cost nothing extra.

Payments aren't live yet. The refund policy will be published before they launch; see Terms → Refunds.

The expiry policy will be announced before payments launch.

In the desktop app, the overlay and the Settings window use your operating system's content protection, so they don't appear in screen shares, recordings or screenshots. Dertos AI can also hide from the Dock and ⌘-Tab, the menu-bar icon is off by default, and the cursor stays a normal arrow over the overlay.

macOS always shows its own microphone and screen-recording indicators, and no app can hide them. System file pickers can't be hidden. In the Chrome web app, a browser tab can't be hidden from a full-screen share, so share a single tab or window instead. We don't claim Dertos AI is invisible to everything.

On your computer. Settings, history, documents and screenshots stay local, and API keys are encrypted with your macOS Keychain. The only network traffic is your own requests to the providers you choose; offline mode sends nothing.

Use Dertos where AI assistance is allowed. Check your interview's or meeting's rules first. You are responsible for following them.

Anthropic Claude, OpenAI, Google Gemini (free tier available), Ollama on your own machine, or any OpenAI-compatible endpoint. Transcription uses Deepgram, OpenAI Realtime or offline Whisper.

For cloud models, yes: bring your own key. One OpenAI key covers both answers and transcription. A free Gemini key from Google AI Studio works for answers, and Whisper + Ollama need no key at all.

A macOS app (macOS 13+, Apple Silicon and Intel) and a web app for desktop Chrome, Edge, Arc and Brave. Windows and mobile aren't available yet.

Anything that plays audio on your computer, including Zoom, Google Meet, Microsoft Teams, Webex, CoderPad and HackerRank. The desktop app auto-detects Zoom, Meet, Teams and Webex calls.

Yes. Every shortcut can be rebound, with conflict detection and reset to default.