Image to text
Upload or paste an image and get its text to copy or edit. It reads English and Spanish, and it flags uncertain words so you can check them.
Drop your files hereTap to choose your files
or click to choose · paste with Ctrl+VJPG, PNG, WebP, GIF, BMP, AVIF, and PDF · up to 100 MB each
Processed on your device: nothing is uploaded
The file couldn’t be opened
Files
Preview
Drag on the image to read only a part of it. With no area, the whole image is read.
Empty = all pages.
Options
The first time, a few megabytes of language data are downloaded. After that, they stay saved in your browser.
Enlarges small images, raises the contrast, and inverts dark backgrounds.
No text found
Extracted text
If you change the format or the options above, the text is generated again and your corrections are lost.
Underlined words have low confidence: compare them with the image. Fix the text in the “Text” view.
How to use it
- Drag your image or scanned PDF onto the box, choose it from your device, or paste a screenshot with Ctrl+V.
- Choose the language of the text. If you only need part of the image, select the area you want to read.
- Select “Extract text.” The first time, about 3 MB of language data is downloaded.
- Check the words marked as uncertain, fix what you need, and select “Copy text” or “Download .txt.”
Frequently asked questions
Is my image uploaded to a server?
No. The text recognition runs inside your browser, on your own processor. The image and the text never leave your device, so you can use it with private documents.
How accurate is the text recognition?
With clean printed text it gets the vast majority of words right, including accents and the ñ when you pick Spanish. No OCR is perfect, though: the tool underlines low-confidence words so you can check them before you use the text.
Which images give the best results?
Images with printed text, good contrast, and letters at least 12 px tall: screenshots, scans, and photos of documents that are straight and well lit. Handwriting, crooked photos, stamps, and complex tables are recognized less well.
Can I read only part of an image?
Yes. Select “Choose an area” and drag on the image to mark what you want to read. It helps you skip menus, ads, or columns you do not need. You can also rotate the image if it is sideways.
Does it work with scanned PDFs?
Yes. When you load a PDF, each page is drawn and read as an image, and you can choose which pages to read. If your PDF already has selectable text, “PDF to text” extracts it instantly and without recognition errors.
Why does the first run take longer?
To read the text, the tool has to download the recognition engine and the language data (about 3 MB for English and 2 MB for Spanish). They are downloaded once and kept in your browser, so later runs start right away.
How to get a good result
- Use the right language. “Spanish and English” recognizes both, but it is a little slower and can mix up similar words.
- Better image, better text. For a photo, keep the page straight, without shadows, and with all the type in focus. For a screenshot, take it at full size: do not shrink it.
- Leave image enhancement on. It enlarges small images, raises the contrast, and inverts dark backgrounds, which OCR reads worse.
- Choose the area. Leaving out borders, logos, and columns you do not need avoids meaningless lines in the result.
What the marked words mean
The engine gives each word a confidence percentage. Words below 75% are underlined in the “Uncertain words” view, and those below 50% are highlighted more strongly. They are usually numbers, proper names, acronyms, and words with unclear accent marks.
When another tool fits better
If the PDF has text you can select, use PDF to text. If you need an editable document with headings, lists, and tables, try PDF to Word.
Updated on September 30, 2026