OCR Text Recognition
Online OCR text recognition tool for extracting Chinese and English text from images with progress feedback, copy and TXT download, fully local processing
Upload Image
Drag an image here or click to upload
Supports JPG, PNG, WebP, BMP, GIF, TIFF, AVIF, up to 20MB per file
Image Preview
Recognition Language
Recognition Settings
Language data for the current selection is preloaded on page entry, so the first load may be slower.
All OCR processing runs locally in your browser and does not upload image content.
Recognition Progress
Recognition Result
Upload an image and start OCR to see extracted text here
Result Stats
Characters: 0
Lines: 0
Language: -
Confidence: -
Duration: -
Tips
- Use clear, high-contrast images for better recognition quality.
- For dark text on a light background, enhancement often improves stability.
- For receipts or headings, try Single Text Line mode.
- You can edit the result, then copy it or download it as TXT.
Guide
OCR Text Recognition is powered by Tesseract.js and extracts Chinese or English text from images with page mode controls and optional preprocessing for screenshots, receipts, and document workflows.
Key Features
- Built on the open-source Tesseract.js OCR engine with local in-browser recognition.
- Supports Chinese and English OCR with real-time progress updates.
- Supports Auto, Single Block, Single Line, and Sparse Text modes.
- Supports black and white preprocessing to improve some inputs.
- Supports editing, copy, and TXT download of recognition results.
Use Cases
- Convert screenshots into editable text
- Extract text from receipts and forms
- Digitize documents and study notes
- Quickly capture text from bilingual assets
FAQ
- Why is the first run slower: language data files are downloaded and cached first.
- How to improve results: use clearer images, enable enhancement, and switch page modes.
- Which languages are supported: built-in Simplified Chinese, Traditional Chinese, and English.
- Why Tesseract.js: it is open source, auditable, can run offline, and is privacy-friendly because images stay local.
- What are the trade-offs: accuracy can drop on complex layouts, handwriting, low-quality or noisy images, and speed depends on device performance.
- Are images uploaded: no, OCR runs entirely in your local browser.
Privacy
OCR processing, result generation, and file export are executed locally in your browser. Images and text are not uploaded to the server.