Image to text.

Extract English text from screenshots, photos, and scans.

Stays in your browser.

No uploads. Nothing saved.

Copy text from screenshots and photos privately in your browser

Drop, choose or paste a screenshot, slide or document photo to extract English text locally with WebAssembly OCR. Copy the editable result or download it as a .txt file.

  • Runs Tesseract LSTM WebAssembly locally with zero image uploads
  • Paste screenshots with Ctrl/Cmd+V and auto-enhance dark-mode images
  • Copy editable text or download a plain UTF-8 .txt file

Common questions

Are my images or extracted text uploaded to any server?

No. Recognition runs strictly inside a local WebAssembly worker in your browser. On the first scan, your browser fetches Tulo's self-hosted 2.8 MB English OCR model and WASM binary; your image bytes never leave your device.

Which image formats and languages are supported?

Supports PNG, JPG, WebP, GIF, BMP and AVIF images up to 25 MB (50 MP / 16,383 px per side). Recognition currently targets English printed text.

How can I improve accuracy on dark screenshots or photos?

Keep Auto-enhance enabled to stretch contrast and automatically invert dark-mode screenshots to dark-on-light text. For a single paragraph or short label, switch Page layout to Single text block or Single line; handwritten notes or heavily skewed photos are less reliable with local LSTM OCR.

Does this work well on phones?

Yes. Large camera photos are automatically scaled down to a memory-safe 2,400 px long edge before recognition, and the OCR worker is terminated when you clear or leave the page to release memory.