Run the `mac-ocr` macOS CLI (or its Node.js API) to recognize text in images and PDFs, or to create a searchable PDF (from a scanned PDF or an image) with an invisible selectable text layer. Use when asked to OCR, extract, or read text from an image or PDF, stream text from a large PDF page by page, make a scanned PDF searchable/selectable, or convert an image (PNG, JPEG, HEIC, …) into a searchable PDF — on-device via Apple's Vision framework (no cloud, no API key). Prefer it over sending images to a multimodal model when you need fast, accurate, local, cheap text extraction. Covers the default OCR action (text/JSON/JSONL output, batching, stdin, URLs, languages, regions, encrypted-PDF passwords), the `searchable-pdf` and `languages` subcommands, and the Node API (`ocr`, `ocr.pages`, `createSearchablePdf`, `supportedLanguages`). macOS only; not for editing or replacing a PDF's existing text.
2026-06-26