How It Works

Four simple steps from image to clean, selectable text — everything happens directly on your device.

1

Drop or Paste

Drop images or scanned PDFs onto the page — or just press Ctrl+V to paste a screenshot straight from your clipboard. Nothing is uploaded; your files stay on your computer.

2

Read

Fast text capture runs right in your browser (about a 31 MB one-time download, then it works offline). Scanned PDFs are read page by page — pages that already contain real text are extracted instantly without re-reading.

3

Review

See the image and the extracted text side by side. Lines the reader wasn't sure about are underlined so you know exactly what to double-check. Click any line in the text to see where it came from in the image — and edit anything before exporting.

4

Copy or Export

Copy everything to the clipboard with one click, or download as .txt, .md, or .csv. Batches export as a single ZIP; multi-page PDFs become one stitched document.

100% Secure

  • Your images never leave this device — receipts, IDs, contracts, and medical letters are read on your computer, not on a server
  • Network is only used once, to download the text-reading engine (then it's cached)
  • Disconnect your internet to test — the tool keeps working entirely offline

Supported Inputs

Screenshots & images PNG, JPG, WEBP, BMP, GIF (first frame), HEIC where your browser supports it
Scanned PDFs Multi-page, read page by page — pages with real text are extracted instantly
Clipboard Paste a screenshot anywhere in the app with Ctrl+V (⌘V on Mac)
Batches Many files at once — no file caps, no queues

Frequently Asked Questions

Does it work offline? +
Yes. The text-reading engine downloads once (about 31 MB) and is saved in your browser. After that, the tool works fully offline — disconnect your internet and try it. Your images are never uploaded anywhere at any point.
Is it really free? +
Yes — free with no account, no watermarks, and no page caps. Everything runs on your own device, so there are no server costs to pass on to you.
What about tables? +
Simple grids (like clean receipts) can be copied as CSV using an experimental table detection. Complex tables — merged cells, borderless layouts — are genuinely hard, and we label table output as experimental rather than pretending it's perfect. A dedicated document-structure mode with real table extraction is planned for capable hardware.
Which languages are supported? +
English, most European languages, and Chinese & Japanese all work out of the box. Korean, Russian & Ukrainian, Hindi, and Thai download automatically the first time you use the tool (about 8–14 MB each) — you just tick which languages are in your files, and once downloaded everything works offline.
Can it read handwriting? +
Honestly: handwriting is hit-or-miss. The tool is built for printed text — screenshots, documents, receipts, scans. Neat block letters may work; cursive handwriting will likely need heavy correction. We flag low-confidence results so you know what to double-check.