FAQ
Questions? We've got answers.
Everything people ask before they download — and the ones they ask afterwards.
The basics
A desktop app that runs alongside your live interview. It transcribes the interviewer, reads regions of your screen when you ask it to, and answers across three surfaces — behavioral chat, live coding, and system design. How it works →
All of them. URHIRED AI is a desktop overlay rather than a plugin or a browser extension, so it sits on top of whatever is on screen — Zoom, Meet, Teams, Webex, Chime, CoderPad, HackerRank, CodeSignal, LeetCode. Setup guides per platform →
Yes — Windows 10 and 11, plus macOS 12 or later on both Intel and Apple Silicon. There is no mobile app.
Chat for behavioral rounds, Coding for algorithm problems, and System design for architecture discussions. Each is a separate thread with its own model and response format, and you switch between them with ⌘1 / ⌘2 / ⌘3 mid-interview.
Staying undetectable
No. The overlay is marked as protected content, so screen-capture and screen-share APIs exclude it from the stream. You see it on your display; the shared frame does not contain it. How it works →
Start a meeting with yourself in whichever app the interview uses, share your screen, and check the share preview. Thirty seconds per app and you know exactly what your interviewer sees. Step-by-step →
No. It hides from your screen share, your Dock, your taskbar, and your app switcher — but the process is listed like any other running application. We would rather you know the boundary than assume one that does not exist.
It is not built to. If an assessment is proctored, that is a disclosed condition of the process you agreed to, and defeating it is not something we build for. Our position →
Share a window or your entire screen. Tab sharing captures the page's own render surface, which is a different path from a desktop capture and behaves differently.
Models and features
GPT-5.4, GPT-5.4 Mini, GPT-5.3 Chat, GPT-5.3 Codex, GPT-5.2, GPT-4.1 Mini, Claude Opus 4.6, and Claude Sonnet 4.6 — and you can assign a different one to each surface. Gemini 2.5 Pro and Grok 3 are coming soon. Full catalog →
No. Model access runs through your account server-side. There is no key to paste and no separate provider bill to manage.
Eighteen: C, C#, C++, Dart, Elixir, Go, Java, JavaScript, Kotlin, PHP, Python, R, Ruby, Rust, Scala, SQL, Swift, and TypeScript.
Five ways of shaping every response: Conversational, Hints, Structured, STAR, and Explain. Switch from the control bar mid-round. More →
Yes, and Hints mode is built for exactly that — it nudges you toward the answer instead of handing it over, so you still do the thinking.
Plans and billing
Yes — session minutes every day, all three surfaces, and undetectable mode, with no credit card. Frontier models and per-surface model selection need Pro or Pro+.
Free is $0. Pro is $19/month. Pro+ is $39/month. Full comparison →
Settings → Profile & Plan in the desktop app opens the billing portal. Cancelling stops future renewals and you keep access through the period you already paid for.
Yes, subject to our refund policy. If something genuinely did not work and support could not fix it, write to us.
Privacy and ethics
Interview turns go through our backend to the model providers to generate responses. See the privacy policy for what is stored and for how long.
No. Screenshots are captured only when you press a capture shortcut. Audio is processed for live transcription rather than stored as a recording.
That depends on the process you are in, and it is your call. Some companies now permit AI assistance; many do not. Follow the rules you agreed to, and disclose where disclosure is required. Our position →
Still stuck?
The user guide covers every feature in depth, and support reads everything.