Media
4 items
Description
Screenshot-to-text OCR. 20+ languages. Runs 100% locally — no image or text is ever sent to any server. FUJI OCR lets you select any region on screen and instantly extract the text from it. OCR runs entirely on your device using WebAssembly — no image or text is ever sent to any server. Select a region with Alt+Shift+S or click the toolbar icon. Results appear in the side panel with per-line confidence scores and full capture history. Supports over 20 languages across multiple scripts. Switch models from the popup before capturing. Export results as plain text, Markdown, or PDF. No account needed. Models are downloaded once and cached on your device — works offline after the first use.