Image to Text (OCR)
Extract text from photos, screenshots, scans, and PDFs right in your browser — pick a language, click Extract, and copy or download the result. Nothing is uploaded.
🔒 Nothing is uploaded — OCR runs in your browser. The recognition engine and language data download from a CDN on first use (a few MB), so the first run needs a network connection and may take a few seconds; after that it runs offline. Accuracy is best on clear, high-contrast, upright text.
Understanding the Image to Text (OCR)
Image to Text (OCR) extracts written text from images and PDFs so you can copy, edit, or search it. Upload a photo, screenshot, scan, or PDF (PNG, JPG, WebP, BMP, or .pdf), choose a language, and click Extract — the recognized text appears in a textarea with live character and word counts, ready to copy or download as a .txt file. The OCR engine, Tesseract, supports 100+ languages, and this tool offers 16 of the most common — English, Spanish, French, German, Italian, Portuguese, Dutch, Russian, Polish, Turkish, Vietnamese, Chinese (Simplified), Japanese, Korean, Arabic, and Hindi. It is for students, writers, researchers, and anyone retyping text from a picture. Everything runs entirely in your browser using tesseract.js; your image and the extracted text never leave your device, so it is private by design and free to use.
How it works
The tool uses tesseract.js, a WebAssembly build of the Tesseract OCR engine, loaded lazily in the browser. On click it spins up a worker for your chosen language code (e.g. eng, fra, jpn), then runs recognition. For an image file the worker reads the File directly. For a PDF it lazy-loads pdf.js, renders each page to a canvas at scale 2 for sharpness, runs OCR on each canvas, and joins the page texts with blank lines. A logger callback drives the progress line, and the worker is always terminated in a finally block to free memory. The engine and language data download from a CDN on first use, then cache for offline runs.
Worked example
You photograph a printed receipt and upload the JPG, leaving the language on English. After clicking Extract, the status line shows 'recognizing text — 100%' and the result box fills with the line items and totals, reporting something like '512 chars · 84 words'. You hit Copy text to paste it into a spreadsheet. Next you upload a 3-page scanned PDF: the tool renders and reads each page in turn ('Rendering page 2 of 3…'), then joins all three pages into one block you download as extracted-text.txt.
Tips & common mistakes
- Use clear, high-contrast, upright images — skewed, blurry, or low-light scans hurt accuracy badly.
- Pick the language that matches the document before extracting; the wrong language produces garbled output. The tool offers 16 common languages, from English and Spanish to Chinese, Japanese, Arabic, and Hindi.
- The first run downloads a few MB of engine and language files from a CDN, so it needs a connection and may take several seconds; later runs are faster and work offline.
- Large multi-page PDFs are processed page by page and can take a while — be patient and watch the progress line.
- OCR is great for printed and clean typed text; handwriting and decorative or stylized fonts are recognized poorly.
- Always proofread the result — even good OCR can confuse characters like 0/O, 1/l/I, or rn/m.
Related tools
Frequently Asked Questions
Which languages are supported?
OCR runs on Tesseract, which supports 100+ languages. This tool offers 16 of the most common ones — English (the default), Spanish, French, German, Italian, Portuguese, Dutch, Russian, Polish, Turkish, Vietnamese, Chinese (Simplified), Japanese, Korean, Arabic, and Hindi. Pick the language that matches your image for the best accuracy.
Does it work on photos and scanned PDFs?
Yes. You can upload images (PNG, JPG, WebP, BMP) such as photos and screenshots, or PDF files. For a PDF, each page is rendered and read in turn, and the text from every page is combined into one result.
Is my image uploaded to a server?
No. The OCR runs entirely in your browser, so your image and the extracted text never leave your device. Only the recognition engine and the chosen language data are downloaded from a CDN on first use.