A few weeks ago someone asked in our Discord why they burned through credits twice as fast on desktop. The answer is the pricing model, which differs by platform, and it matters more than most people realize when you’re mid-interview and watching a timer.
Craqly ships three separate interfaces: a native desktop app, a browser-based web app, and a mobile app. They share the same credit pool. That’s where the similarity ends.
What the desktop app actually does that the others can’t
The desktop version is the only one that supports coding interview mode. If you’re walking through a live coding problem and want context-aware suggestions, you need desktop. The web and mobile apps don’t have that mode.
More importantly, desktop has stealth mode. Toggle it with Ctrl+B (or Cmd+B on Mac) and the app becomes invisible to screen-recording and screen-sharing software. I’m honestly not 100% sure how solid this is across every possible screen-share setup, but the core mechanism works reliably for the major conferencing tools. If your interviewer is watching your screen, this matters a lot.
Desktop also lets you pick your underlying AI model. GPT-4, Claude, Gemini. The web and mobile apps don’t expose that choice.
The cost: 1 credit per minute.
The credit math favors web and mobile
Web and mobile both run at 1 credit per 2 minutes. That’s half the burn rate of desktop. For a 47-minute behavioral interview where you’re not sharing your screen, the difference adds up to roughly 23 credits saved per session.
The web app supports live interview mode with screen analysis. It reads what’s on screen and feeds that context to the AI. The mobile app works differently: it captures audio through your phone’s microphone, so you can use it as a second device sitting off-camera while your laptop runs the actual interview. That’s a useful setup people underestimate.
When to use which
Here’s the short version:
- Coding interview (any company): Desktop, full stop. You need coding mode and you probably want stealth on.
- Behavioral interview, no screen share: Web app. Better credit efficiency, same AI quality.
- Interview on a locked-down work machine where you can’t install software: Web app. Nothing to install.
- You want a second device running quietly: Mobile app. Put your phone face-down near your keyboard, mic picks up the conversation.
The Stack Overflow Developer Survey 2024 found that 62% of developers do technical interviews that include a live coding component. For that group, desktop is the right choice. For everyone else, the web app is usually the better tradeoff.
Resume-based personalization is desktop-only for now
Desktop lets you upload your resume so the AI tailors its suggestions to your actual background. If an interviewer asks about a project and you want the response grounded in your specific experience rather than generic advice, that context comes from the resume upload. Web and mobile don’t have this feature yet, though it’s on the roadmap.
This is probably the strongest argument for desktop even in non-coding interviews, at least until the web app catches up.
One thing that surprises people
The credit pool is unified across all three platforms. This surprises a lot of users who expect platform-specific credit buckets. You buy once, use anywhere. Switch mid-interview if you need to. The session context doesn’t transfer automatically, but credits do.
Which platform are you on most often? The answer probably says more about your interview mix than anything else.