Extract text from an image
Pick the image files, get their text. Up to 20 at a time, each with its own text, language and confidence, and one download for the lot.
Drop images or PDFs here, or click to choose
PNG, JPG, WebP, GIF, BMP, TIFF, PDF
What you get
- JPG, PNG, WebP, GIF, BMP and TIFF, identified by their bytes, not their name
- Paragraphs and line breaks kept from the layout
- The language of the text, detected
- A confidence figure per image
- One .txt or .md file for the whole upload
How it works
- Upload. Your files go to the server in parts; each is identified by its bytes, not its name.
- Read. A document OCR engine finds the blocks, paragraphs and lines on each page and detects the language. Files are read one after another.
- Use. Copy the text per file, or download everything as one .txt or .md. Then it is gone: see below.
Limits
Up to 20 files per upload, 7 MB per file and 50 MB in all. PDFs and TIFFs of up to 5 pages; a longer file is refused with a message.
Privacy
Uploads are kept only while you work with the result and are deleted automatically after 1 hour. Nothing is stored beyond that: no account, no history, no copies of your files or their text.