Is my image uploaded for OCR?

No. The OCR engine (Tesseract, compiled to WebAssembly) downloads to your browser once and runs there — your image stays on your device.

On clean screenshots and printed text, very. Accuracy drops with blur, skew, handwriting, and low contrast — a sharp, straight photo makes a big difference.

Does it read handwriting?

Not reliably — OCR is built for printed and rendered text. Neat block capitals sometimes work; cursive rarely does.

Which languages does it support?

English works out of the box — recognition quality is best on Latin-script text.

Can I make a scanned PDF searchable instead?

Yes — that's the Scan tool: it stacks photographed pages into a PDF and can embed an invisible OCR text layer, so the PDF becomes selectable and searchable.

Image to text, on your device.

Drop a screenshot or photo and the text is recognized locally — the OCR engine loads into your browser; your image never leaves it.

Extract text from an image in your browser — screenshots of error messages, photos of book pages, whiteboard shots, receipts. GhostImage runs OCR (optical character recognition) locally on your device and hands you the text, ready to copy or download as a .txt file. The OCR engine downloads once to your browser, but your image never leaves it — which matters when the screenshot contains a stack trace from work, an address, or anything else you'd rather not feed to a stranger's server.

How it works

1
Drop your image
Drag in a screenshot, photo, or scan. It opens in the GhostImage studio with Extract text pre-selected.
2
OCR runs on your device
The recognition engine loads into your browser and reads the text locally — the image itself is never sent anywhere.
3
Copy or download the text
The recognized text is yours to copy or save as .txt — no accounts, no per-page pricing.

Frequently asked questions

Is my image uploaded for OCR?
No. The OCR engine (Tesseract, compiled to WebAssembly) downloads to your browser once and runs there — your image stays on your device.
How accurate is it?
On clean screenshots and printed text, very. Accuracy drops with blur, skew, handwriting, and low contrast — a sharp, straight photo makes a big difference.
Does it read handwriting?
Not reliably — OCR is built for printed and rendered text. Neat block capitals sometimes work; cursive rarely does.
Which languages does it support?
English works out of the box — recognition quality is best on Latin-script text.
Can I make a scanned PDF searchable instead?
Yes — that's the Scan tool: it stacks photographed pages into a PDF and can embed an invisible OCR text layer, so the PDF becomes selectable and searchable.

Related image tools

Document scanner

Photograph a document and get a clean, straightened, searchable PDF.

Crop image

Drag a box over the part you want to keep.

Compress image

Shrink JPG, PNG, WebP, AVIF, or HEIC with a quality slider.

PDF to Word

Rebuild a PDF as an editable .docx — text and structure transfer, locally.

Every GhostImage tool runs entirely in your browser. Your file is never uploaded — there's no upload endpoint to send it to.