Convert a photo or screenshot to text — free, without uploading it
You have a photo of a page, a screenshot of an error message, or a scan of a receipt — and you need the text, not the picture. Most "image to text" websites work by uploading your image to their server. This one doesn't: the OCR engine runs inside your browser, and the image never leaves your device.
Drop in a photo or screenshot — PNG, JPEG, WebP or BMP.
Wait a few seconds while the OCR engine (bundled inside the page) reads it.
Copy the text to your clipboard or download it as a .txt file.
The first scan loads the OCR model (a few seconds); every scan after that is fast. Everything happens locally — you can disconnect from the internet and it still works.
When is OCR useful?
Screenshots: error messages, code, chat windows, tables — grab the text instead of retyping it.
Photos of documents: receipts, business cards, whiteboards, printed pages you'd rather not retype.
Scanned paperwork: get a searchable, editable copy of a printed letter or form.
Quotes and references: copy a paragraph from a photo of a book page instead of typing it out.
What OCR does well — and where it struggles
OCR is excellent on printed text: clean scans of books, printouts and screenshots are read very reliably. It's weaker on handwriting — cursive especially — and on blurry, angled or low-light photos. Two habits fix most misses: shoot straight-on with even lighting, and crop to the text you actually need. And proofread anything important; OCR confidence is high but not perfect.
Why "upload your image" is the wrong default
Images people OCR are often sensitive: contracts, invoices, medical letters, screenshots of private conversations. On an upload-based service, each one makes a round trip to someone else's server. Here the engine ships inside the page itself — the same technology desktop OCR apps use, compiled to WebAssembly. Open your browser's network inspector and watch it make zero requests while scanning. That's the proof.
scan.png — reading on this device: 412 words recognized language: English ✓ text copied — never uploaded, 0 bytes sent
What an OCR run looks like (illustration — demo image, never real data):
Like the tool? Own it.
The free page scans unlimited single images. The €9 OCR Scanner download adds batch scanning of whole folders and the same tool as a single file — no limits, offline forever, on any computer you own. One-time payment — no subscription.
No. The OCR engine is bundled inside the page — there is no server behind it. Load it, disconnect from the internet, and it still works. The network inspector will show zero requests during a scan.
Which languages can it read?
The free page ships with the English model. Other languages are planned; the downloaded version adds batch folder scanning in the same single file.
Is the text editable?
Yes — output is plain text you can edit in place, copy to the clipboard, or download as a .txt file.
What's the difference between this free page and the paid download?
This page scans unlimited single images, one at a time. The €9 download adds batch scanning of whole folders and runs offline forever as a single HTML file on any computer.
Untracked builds single-file tools: no installation, no account, no cloud, no subscription — and no tracking on this page. This guide lives on tools.untracked.tools; every claim on it is about tools you can open and verify yourself.
We use one first-party analytics tool (PostHog, hosted in the EU) to understand which pages and tools are useful. It never sees your files, form contents or search terms, and session recording is off. You can change your mind at any time.