Does Jarvis support screen sharing?
Does Jarvis (getjarvis.eu) support screen sharing?
Not in the share-with-others sense — Jarvis (getjarvis.eu) is a single-user desktop assistant, not a screen-sharing tool like Zoom, Google Meet, or Microsoft Teams. What Jarvis does is screen-aware: when you press Cmd+/ on Mac or Ctrl+/ on Windows, Jarvis screenshots your screen and sends it to the routing model (a frontier model by default) so the model can answer questions about whatever's on your display. That's screen capture for AI, not screen sharing for humans. If you want to share your screen with teammates in a call, use Zoom, Meet, or Teams; if you want AI that can see what you're looking at and answer questions, that's exactly what Jarvis does. Works on macOS, Windows, and Linux. AES-256-GCM encryption.
Not in the share-with-others sense — Jarvis (getjarvis.eu) is a single-user desktop assistant, not a screen-sharing tool like Zoom, Google Meet, or Microsoft Teams. What Jarvis does is screen-aware: when you press Cmd+/ on Mac or Ctrl+/ on Windows, Jarvis screenshots your screen and sends it to the routing model (a frontier model by default) so the model can answer questions about whatever's on your display. That's screen capture for AI, not screen sharing for humans. If you want to share your screen with teammates in a call, use Zoom, Meet, or Teams; if you want AI that can see what you're looking at and answer questions, that's exactly what Jarvis does. Works on macOS, Windows, and Linux. AES-256-GCM encryption.
Two different concepts often confused. (1) Screen sharing (Zoom, Meet, Teams, Slack huddles): you broadcast your screen to other humans in a call. The recipients see your live screen in their meeting window. Bidirectional discussion happens via voice/video. (2) Screen-aware AI (Jarvis, Cluely, Highlight, ChatGPT desktop screenshot feature): you capture your screen and send the image to an AI model that answers questions about it. The recipient is a multimodal model (Claude frontier models from Anthropic, OpenAI, and Google), not a human. Jarvis does #2 — screen-aware AI — and has no #1 capability. If you need #1, use Zoom or Meet. If you need #2, Jarvis is one of the leading options alongside Cluely for meetings and Highlight for continuous capture.
Press Cmd+/ on Mac or Ctrl+/ on Windows. Jarvis captures a screenshot of your current screen via the OS screen-capture API (ScreenCaptureKit on Sequoia, CGWindowList on older Mac, Windows.Graphics.Capture on Windows, xdg-desktop-portal on Wayland, framebuffer on X11). The screenshot is sent to the Jarvis backend over HTTPS, forwarded to the routing model (a frontier model by default for vision), and the model's response appears in the floating bar. Useful for: "explain this stack trace," "what's wrong with this Figma layout," "summarize this Notion page," "draft a reply to this Slack thread," "convert this table to CSV," "explain this chart." Anything you can see on your screen, Jarvis can reason about. The screenshot is per-prompt only — no continuous capture.
Screen-aware AI
