Text
Image to Text (OCR)
Your image is never uploaded — text extraction runs entirely on your device. A private way to OCR receipts, IDs, and documents.
Most online OCR tools send your photo to a server before giving you anything back — fine for a random screenshot, less fine for a receipt with your card's last four digits, a passport page, or a signed contract. This tool runs the entire recognition pipeline locally, inside a background Web Worker powered by Tesseract.js, so the image data never leaves your browser tab. Drop in a photo or screenshot, pick the language, and get back editable, searchable text you can copy directly or download as a plain .txt or formatted .docx file.
Related Tools
How to Use
- 1
Drop a photo or screenshot onto the upload area, or click to browse.
- 2
Choose the language of the text in your image from the dropdown.
- 3
Click "Extract Text" — recognition runs locally in a background Web Worker and typically takes a few seconds.
- 4
Copy the extracted text directly, or download it as a .txt or .docx file.
Frequently Asked Questions
Is my image uploaded anywhere?
No. The image is decoded and processed entirely inside your browser using a Web Worker running Tesseract.js — it never leaves your device or touches a server. That matters if you're scanning a receipt, an ID, a contract, or anything else you'd rather keep private.
How accurate is the text extraction?
For clear, well-lit, high-resolution photos of printed text, accuracy typically exceeds 95%. Blurry photos, low light, extreme angles, or low resolution reduce accuracy — a sharper photo cropped closer to the text usually helps most.
Can it read text from a photo of a receipt or document?
Yes — it works well on printed receipts, invoices, book pages, signs, screenshots, and scanned documents. Photographed receipts with faded or curled thermal paper can be lower accuracy; flattening the receipt and using good lighting improves results.
Does it work on handwriting?
Tesseract is optimized for printed text. Very neat, large handwriting may partially recognize, but accuracy on handwriting is generally much lower than on printed text.
Which languages are supported?
English, French, German, Spanish, Portuguese, Italian, Simplified Chinese, Japanese, and Arabic. Select the language of the text in your image before clicking Extract Text.
Why does it take a moment before it starts?
The language model is about 4 MB and downloads once on first use, then is cached in your browser for instant use afterward. Recognition itself runs in a background Web Worker so the page stays responsive while it works.