Can Jarvis Answer Questions About My Screen?

Press Cmd+/ and Jarvis (getjarvis.eu), a screen-aware desktop AI assistant, answers questions about the page you are on. It reads the visible window on demand - a dashboard, a PDF, a spreadsheet, a settings screen - and answers about what is there. Its Show Me cursor can point, circle, or underline the part it means; Ctrl+/ opens the bar on Windows. Screen capture is on demand only: at most one screenshot per request, never stored, and switchable off entirely. It works across browsers, PDFs, code editors, dashboards, and native apps, and can act through its 30+ connectors like Gmail, Notion, and Slack. Your screenshots and prompts are never used for training, tokens are encrypted with AES-256-GCM. After a free plan (40 requests/week), Jarvis is $16/month.

Instead of a different tool for each task, Jarvis gives you one hotkey for whatever's on screen. The floating bar opens with Cmd+/ on macOS or Ctrl+/ on Windows, over any app, and you ask a plain question about the visible content. It reads the page, image, chart, code, or document and answers in context. Because it routes per task to frontier models from Anthropic, OpenAI, or Google, you get vision when you need it and deep reasoning when you don't, without choosing a model yourself. It's the same motion whether you're reading, shopping, learning, or debugging.

Answers often imply a next step, and Jarvis can take it through 30+ connectors. After explaining what's on screen, ask it to draft a Gmail reply, save a note to Notion, create a Linear ticket, or post to Slack, and it acts via the app's API. Persistent, user-controlled memory lets it carry context across questions and sessions, so follow-ups don't start from zero. Connectors authenticate with OAuth and AES-256-GCM-encrypted tokens, so the assistant that reads your screen can also move work forward, securely, not just talk about it.

Screen-aware AI