use case · ask ai about your screen
Half of what you would ask an AI is about something you are looking at: an error, a chart, a clause, a number that cannot be right. Do not translate it into a paragraph of description. Crop the thing, ask while you circle it, and paste — the screenshot carries what you see, your words carry what you want to know.
No — ChatGPT, Claude, Gemini and Copilot see only what you choose to hand them. That is good for privacy, but it leaves you doing the describing. CroppickVoice turns the hand-over into one gesture: crop the exact part of the screen, mark it up while you ask your question out loud, and paste.
The screen dims and recording starts. Drag over the error, the chart, the clause — only that region is captured.
Circle the number, arrow the button, box the sentence — undo is a click away — and say what you want to know as you draw.
One paste: the annotated screenshot and your question, already written out. Read the question over, then send.
Any chat that accepts a pasted image — ChatGPT, Claude, Gemini, Copilot — gets the picture and your words together. Text-only places like terminals get your words plus the image's file location instead, which is what a tool like Claude Code needs. Got a path where you wanted the picture? Open History from the button and click the entry: that second delivery often lands the image where the first paste could not. How delivery adapts per app →
Once the screenshot has done its job, the rest is conversation: press Ctrl ` and your reply lands in the chat box as editable text. That half has pages of its own — dictating to ChatGPT and dictating to Claude.
No — ChatGPT, Claude, Gemini and Copilot see only what you choose to hand them. To ask about something on your screen, crop that part, mark it up while you ask out loud, and paste: the annotated screenshot and your transcribed question arrive in the chat together.
No. Only the region you dragged over, with your markup on it. The rest of the screen never leaves your computer — screenshots are saved to a local folder, and the app tracks nothing.
No. Crop & Speak puts the annotated image and your transcribed question on the clipboard together; one paste in the chat delivers both.
Yes. Alt+1 captures and annotates without recording; type your question next to the pasted image instead.
Screenshots stay in a local folder and speech becomes text on your device — the AI sees what you paste, and only that. Related: bug reports with annotated screenshots, dictation for students and all use cases.