About the Image to Text (OCR)
Optical Character Recognition turns a picture of text — a photo of a document, a screenshot, a scanned page — into real, selectable, editable text. Instead of retyping, you upload the image and let the OCR engine detect the characters, including multiple languages. This tool uses a WebAssembly build of the Tesseract engine, so the entire process happens locally on your device.
After extraction you can edit the recognized text to fix any mistakes, then export it in the format you need: a .docx Word file for further editing, a PDF for sharing, or a .txt file for maximum compatibility. Because everything runs client-side, it is safe to process private documents, IDs or contracts — nothing is sent to a server (only the small language model is downloaded once and cached).
OCR accuracy depends on image quality: sharp, high-contrast, straight images produce the best results, while blurry, skewed or low-resolution photos are harder. For scanned PDFs, extract the pages as images first, or use this tool on a screenshot of each page.
How to use the Image to Text (OCR)
- 1Upload an image containing text (PNG, JPG, WebP, BMP…).
- 2Choose the language that matches the text (e.g. English or Tiếng Việt).
- 3Click “Extract text” and wait for recognition to finish.
- 4Review and edit the recognized text if needed.
- 5Download it as Word (.docx), PDF, or TXT.
Key benefits
- Runs in your browser — images stay private.
- Supports many languages, including Vietnamese.
- Editable result before export.
- Export to Word, PDF or plain text.
Real-world examples
Digitise a document
Turn a photo of a printed page into an editable Word file.
Copy text from a screenshot
Extract text you can't select from an image.
Capture a receipt
Pull the text out of a receipt or invoice photo.
Frequently asked questions
Is my image uploaded anywhere?+
No. OCR runs entirely in your browser via WebAssembly. Only the language model is downloaded (once) from a CDN; your image never leaves your device.
Which languages are supported?+
Many, including English, Vietnamese, French, German, Spanish, Japanese and Chinese. Pick the one matching your text, or the combined English + Vietnamese option.
Why isn't the text accurate?+
OCR works best on clear, high-contrast, upright images. Blurry, rotated or low-resolution photos reduce accuracy. You can edit the result before exporting.
Will the PDF show Vietnamese characters correctly?+
For non-Latin scripts and Vietnamese diacritics, the Word (.docx) and TXT exports are most reliable; the built-in PDF font is optimised for Latin characters.