Web app Personal product
QuickImageToText: Free Online OCR Tool
My own OCR web app, built with JavaScript and Tesseract.js, that turns images and PDFs into editable text in the browser, with no upload, in 8 languages.
Problem, solution and impact
Problem
Copying text out of images is slow, and many OCR tools add sign ups, usage limits, credits or a clumsy import, which hurts most when you have many images.
Solution
I built a browser OCR tool on Tesseract.js: images come in by file, drag and drop, paste or URL, and text is extracted from up to 20 images at once in 19 languages.
Impact
- Anyone can pull text from screenshots, scans and photos with no account and no credits
- Batches of up to 20 images save students and office workers from copying one image at a time
- Four ways to import, including paste from the clipboard, remove the usual upload friction
- The site is available in 8 languages, so people can use it in their own language
Overview
QuickImageToText extracts text from photos, screenshots, scans and handwritten notes. Recognition runs client-side with Tesseract.js, so images never leave the user's device and nothing is stored on a server.
I designed, built and launched it as a personal product, and I handle its SEO myself: localised URLs and hreflang in eight languages, structured data, and content written for both search engines and AI answer engines. It is live and grows through organic search.
After launch I redesigned the whole site for a calmer, faster experience, reworked the extraction pipeline so results come back sooner, and added a blog backed by Supabase. Posts are written in a separate admin app with a drag-and-drop section builder, so anyone can publish without touching code, much like WordPress or Shopify.
Key features
- Client-side OCR: images are never uploaded
- Text recognition in 19 languages, with automatic writing-system detection
- Batch processing of up to 20 images
- PDF-to-text and document OCR tools
- Blog powered by Supabase
- Separate admin app with a drag-and-drop section builder for non-technical editors
- Light and dark themes
- Site localised into 8 languages with hreflang
Articles about this project
-
18 min read
How Tesseract OCR Works in the Browser and on Your Device
How Tesseract OCR turns images into text, how to run it in the browser or on your own computer, with JavaScript, TypeScript, Node, PHP code and real fixes.
Redesigned in 2026
The redesign
The first version worked, but it looked like a template: gradient headings, a different colour on every button and a long, centred layout. I rebuilt the interface around the task itself and made the OCR faster at the same time.
- Cleaner layout: a left-aligned headline and every upload option grouped in one card
- One accent colour, with matching light and dark themes
- Automatic detection of the writing system, plus up to 3 languages for mixed text
- Black-and-white preprocessing option for faint or uneven scans
- Faster, smoother text extraction
- New blog on Supabase, managed from a drag-and-drop admin app
Drag to compare the original design with the redesign