Image to text with OCR in 10 languages
Extract text from scanned images and photos.
To convert an image to text, upload a picture, choose the language of the writing and read the extracted text on the results page. The tool runs Tesseract OCR on one image at a time, in one language you select: English, Spanish, French, German, Portuguese, Simplified Chinese, Hindi, Arabic, Japanese or Russian. Accuracy depends on how sharp and well lit the picture is, so expect to proofread the output.
How to extract text from an image
Choose a JPG, JPEG, PNG, WEBP, GIF or BMP file up to 20 MiB. Set the text language, which defaults to the language of the page you are viewing, then submit. The text appears on the page and you can copy it. If nothing is recognised, you see a message that no text could be extracted.
Pick the right language
OCR works with one language per run, and the language must match the writing in the picture. An English setting applied to a French receipt will misread the accents, and there is no automatic detection or fallback. Mixed-language images are read in the language you picked, so the other language comes out unreliable.
Getting a better result
Use a straight-on, sharp image with dark text on a plain light background, at a reasonably high resolution. Skewed photos, glare, shadows, decorative fonts and low contrast all produce errors. Handwriting is generally not recognised. Tables and multi-column layouts come back as plain text lines in the order the engine reads them, so check the order.
Scans and other file types
This tool reads images only. HEIC photos from an iPhone need converting first with the HEIC Converter. To make a scanned PDF searchable, use Searchable PDF, which adds a text layer to the original pages. The Image Converter can change an unsupported image format to PNG or JPG before OCR.
Frequently asked questions
Which languages does the OCR support?
English, Spanish, French, German, Portuguese, Simplified Chinese, Hindi, Arabic, Japanese and Russian. You choose one language per image, and there is no automatic detection. Other languages, including Traditional Chinese, are not offered.
How accurate is image to text OCR?
On clear, high-contrast printed text it is usually good, but never perfect. Blur, glare, small type and busy backgrounds lower accuracy. Always proofread numbers, names and addresses before relying on the result.
Can it read handwriting?
Generally not. The engine is designed for printed text and handwriting usually comes out garbled or empty. Neat block capitals on a clean background may work better, but do not count on it.
What image types and sizes are accepted?
JPG, JPEG, PNG, WEBP, GIF and BMP files up to 20 MiB, one image at a time. HEIC and PDF are not accepted here. Convert HEIC with the HEIC Converter, or use Searchable PDF for scanned PDFs.
Why did the OCR return nothing?
No readable text was found. Common causes are a wrong language setting, a very blurry or low-resolution picture, or text that is too small or low in contrast. Try a sharper image or a different language.
Related tools
Word to PDF
Convert Word documents into polished PDFs.
Use tool →Excel to PDF
Turn spreadsheets into shareable PDF reports.
Use tool →CSV to Excel
Preview delimited text and create an XLSX workbook while preserving text and leading zeros.
Use tool →Excel to CSV
Preview one XLSX worksheet and export its stored values as delimited text.
Use tool →PowerPoint to PDF
Convert a self-contained PPTX presentation into PDF using an isolated Office renderer.
Use tool →