A macOS OCR keyboard shortcut: extract text from any screenshot

Some text just won't let you select it: a screenshot a colleague sent, a frame paused in a video, an error dialog, a scanned PDF, a slide in a screen-share. You can see the words; you just can't copy them. OCR (optical character recognition) turns those pixels back into selectable text, and the fastest way to use it is a single keyboard shortcut.

Here's how to set that up on macOS with Khint, so capturing text is one gesture: shortcut, drag a box, paste.

The one-shortcut flow

1. 1
   Press Cmd+Shift+KThe Khint palette opens over whatever you're looking at, no need to switch apps first.
2. 2
   Pick 'Extract text'Under Capture, choose Extract text. Khint hands the capture to the macOS screenshot tool.
3. 3
   Drag over the textDraw a box around the words you want. Release to capture.
4. 4
   It's on your clipboardThe recognized text is copied automatically, with a small confirmation showing a preview. Paste it anywhere.

When you want more than raw text: Capture & ask

Sometimes you don't just want the text, you want to do something with it. The palette's Capture & askrow captures a region and opens a short chat about it: "summarize this table," "what's the error here and how do I fix it," "turn this slide into bullet points." It's the same capture gesture with a question attached.

How this differs from macOS Live Text

Apple's Live Text is great when it's available: hover over an image in Photos or Preview and you can often select the text. The gaps it leaves are exactly where a shortcut helps: content in apps that don't support Live Text, video frames, screen-shares, and anything where you'd otherwise screenshot-then-fiddle. A global shortcut works the same way everywhere, and it copies straight to your clipboard instead of needing a manual select-and-copy.

Good things to capture

• An error message in a dialog you can't select or copy.
• A code snippet or command shown in a screen-share or video tutorial.
• A table or figure inside a PDF or an image attachment.
• Text on a slide during a meeting you want to act on later.
• A receipt, a label, or anything you photographed instead of typed.

Because OCR runs from the same palette as your AI Actions, you can go straight from captured text to a transformation: extract a chunk of notes, then turn it into a Jira ticket without ever retyping it.