liminfoOCR Tool
Free web tool: OCR Tool
Drag an image file here or
Choose filePNG, JPG, BMP, WebP, etc. (max 20MB)
About OCR Tool
The OCR Tool is a free, browser-based optical character recognition utility powered by Tesseract.js — the industry-standard open-source OCR engine originally developed by HP and now maintained by Google. It recognizes printed text in images and converts it into editable, copyable plain text entirely within your browser. No file is ever uploaded to a server.
Web developers, researchers, students, administrative workers, and accessibility advocates rely on OCR to digitize scanned documents, extract text from screenshots, process photos of whiteboards or printed forms, and make image-based content searchable. This tool supports PNG, JPG, BMP, and WebP image files up to 20 MB, covering virtually every common image format.
Under the hood, the tool dynamically imports the Tesseract.js library on demand, loads the selected language's trained neural-network data, and runs the full recognition pipeline in the browser. A progress bar tracks each stage — core loading, language data loading, API initialization, and text recognition — giving you clear feedback as the engine works through your image.
Key Features
- Multi-language OCR: English, Korean, Japanese, Simplified Chinese, Traditional Chinese, and English+Korean mixed mode
- Supports PNG, JPG, BMP, and WebP image formats up to 20 MB per file
- Live progress indicator showing Tesseract.js initialization stages and recognition percentage
- Editable result textarea — manually correct recognized text before copying or downloading
- One-click copy to clipboard and plain-text .txt file download named after the original image file
- Image preview before recognition so you can confirm the correct file was selected
- 100% client-side Tesseract.js processing — no image data ever sent to any server
- Dark mode support and mobile-friendly drag-and-drop file zone
Frequently Asked Questions
What is OCR and how does it work?
OCR (Optical Character Recognition) is the process of analyzing pixel patterns in an image to identify characters and convert them into machine-readable text. This tool uses Tesseract.js, a WebAssembly port of the Tesseract 4 engine, which employs an LSTM (Long Short-Term Memory) neural network trained on large multilingual datasets to recognize characters from image data.
Which languages does this OCR tool support?
The tool supports English, Korean, Japanese, Simplified Chinese, Traditional Chinese, and a combined English+Korean mode for documents containing both languages. Select the recognition language before clicking Start Text Recognition to load the appropriate language model. Choosing the correct language significantly improves accuracy.
What image formats and sizes are supported?
Any image format that your browser can display is accepted, including PNG, JPG/JPEG, BMP, WebP, and GIF. The maximum file size is 20 MB. For best results, use high-resolution images (at least 300 DPI for printed text) with good contrast between text and background.
Why does recognition take a while on first use?
On first recognition, Tesseract.js must download the Tesseract core (WebAssembly binary) and the trained language data file for your selected language. These are cached by your browser after the initial download, so subsequent uses of the same language are much faster. The progress bar shows each loading stage.
Is my image uploaded to any server?
No. All OCR processing happens entirely within your browser using Tesseract.js compiled to WebAssembly. Your image is read locally via the FileReader API and processed in memory. No image data, recognized text, or metadata is ever transmitted to any server.
Can I edit the recognized text after OCR?
Yes. The recognized text appears in an editable textarea, allowing you to correct mistakes before copying or downloading. Tesseract may occasionally misread certain fonts, handwriting, or low-contrast text, and the editable output lets you fix those errors directly.
What affects OCR accuracy?
Image quality is the most important factor. Higher resolution (300+ DPI), clean backgrounds, horizontal text orientation, standard fonts, and strong contrast all improve accuracy. Skewed, rotated, low-resolution, or heavily stylized text will reduce accuracy. For handwritten text, accuracy varies significantly by writing style.
How do I download the recognized text?
After recognition completes, click the Download Text button. The tool creates a .txt file named after your original image file (for example, an image named "scan.png" produces "ocr_scan.txt") and triggers a download through the browser's built-in download mechanism without involving any server.