Pictotext

Extract text from any image on the web or any screen region using local Tesseract.js OCR. No server. No API key. Private by design.

Pictotext extracts text from any image on your screen — instantly and privately. Powered by Tesseract.js running entirely in your browser, no data ever leaves your device. No server. No API key. No account required. HOW TO CAPTURE • Select Region — drag a crosshair over any part of the screen • Screenshot — capture the visible tab in one click • Full Page — stitch an entire webpage into one capture • From Image — upload a file or paste a URL • Clipboard — Ctrl+V an image directly from your clipboard • Drag & drop — drop an image file onto the popup WHAT YOU CAN DO WITH THE RESULT • Edit the extracted text before copying • Find & replace to fix common OCR mistakes • Text-to-speech — have the result read aloud • Search — copy and open a Google search in one click • Save the capture as PNG or PDF • See confidence scores per word and filter low-confidence results MORE FEATURES • Annotate before extracting — draw, highlight, or crop the image first • Repeat last region — re-capture the same area in one click, great for live content • Batch processing — queue multiple images and process them in sequence • History panel — last 20 extractions saved locally so you can go back without re-scanning • Multi-language OCR with downloadable language packs PRIVACY Everything runs locally using WebAssembly. No external API calls, no telemetry, no subscription. Your images and text never leave your browser. WORKS GREAT FOR Copying text from locked PDFs, extracting quotes from social media images, grabbing error messages from screenshots, reading any image-based text aloud, and pulling data from receipts, forms, or scanned documents.