Apple introduced Live Text as a system-level feature that recognizes text in images. In theory, it sounds like it solves the "can't select text in images" problem. In practice, it works in a narrow set of scenarios and silently does nothing in many situations where you actually need it.
Where Live Text Falls Short
Live Text works inside Photos, Preview, Safari (on some images), and Quick Look. That's roughly where its reach ends. Here are the common situations where it can't help:
Video frames. Live Text does not operate on video content. If text appears in a YouTube video, a Zoom screen share, a QuickTime recording, or any streaming player, Live Text won't recognize it. You're back to manual transcription.
App interfaces and dialogs. Text rendered as labels in app UIs, error dialogs, menu items, and custom controls is invisible to Live Text. It only activates on embedded images, not on arbitrary screen content.
Web apps with canvas or SVG text. Many modern web apps render text using canvas elements or SVG — dashboards, data tools, design apps. Live Text doesn't reach into these rendering contexts.
Non-selectable PDFs. While Preview can sometimes recognize text in image-based PDFs, the results are inconsistent and depend on scan quality, page complexity, and PDF structure. Many scanned documents simply show no Live Text interaction at all.
Copy-protected content. Live Text respects DRM and copy restrictions. If an app or document has disabled text selection, Live Text follows suit.
No capture history. Live Text has no memory. Every recognition is ephemeral. If you close the image or navigate away, the recognized text is gone. There's no log or history of what you've extracted.
Screen-Level OCR Without These Limits
Optic works at the screen level rather than the file or image level. It reads characters from whatever is visible on your display — any app, any source, any context. Video frames, dialog boxes, web app canvases, locked PDFs, remote desktop sessions — if you can see text, Optic can read it.
History and Text-to-Speech
Unlike Live Text, Optic saves every capture in a persistent history accessible from the menu bar. It also includes text-to-speech, letting you hear captured text read aloud — useful for proofreading or accessibility.
QR Code Scanning
Live Text can read QR codes in some image contexts, but not from video, screen shares, or arbitrary app windows. Optic scans QR codes from anywhere on your screen and decodes them instantly.