Image to text converter – free AI OCR in your browser

How to convert an image to text

1. Add your image

   Drop one or more images on the box, pick them with Choose images, or copy a picture and press Ctrl+V (Cmd+V on a Mac). On a phone, Take a photo opens the camera.

2. Set the language

   Choose English, German, French or Spanish. For photos or tiny text, switch on Boost contrast or Enlarge small images. To read only one part, drag across the picture.

3. Read, check and copy

   Press Read text. The result appears in a box you can edit: correct it, join broken lines if you like, then copy it or download a .txt file.

The first run loads the text engine (about 4 MB) and the language file (0.7–3 MB) from this website. After that the language stays in your browser, and the next image starts much sooner.

Text from photos, screenshots, receipts and slides

Anything with printed or on-screen text is worth a try. A few everyday cases:

• A photo of a sign, a menu or a notice board: read it and paste the opening hours or the dish names into a message.
• A screenshot of a PDF page: when a PDF is only a scan and the text cannot be selected, a screenshot of the page turns it into text here.
• A slide from a lecture or a video call: keep the bullet points as notes instead of typing them.
• Receipts, labels and serial numbers: copy an order number, an ingredient list or the long Wi‑Fi password on a router.
• An old letter or a printed page: get text you can edit, search and quote.

Printed text works best. Neat handwriting is read only sometimes and quick notes rarely; for those, typing is usually faster.

JPG to text file: which images work?

The AI Text Scanner opens the image formats your browser can show: JPG, PNG, WebP, BMP and GIF (the first frame of an animation). HEIC photos from an iPhone work only in browsers that can open HEIC; if one does not open, save it as JPG or take a screenshot of it.

Download .txt saves a plain UTF‑8 text file named after your image, so receipt.jpg becomes receipt.txt.

Why is the text wrong or missing?

Text recognition reads shapes, so the picture decides most of the result. When words come out wrong, one of these usually helps:

• Light and focus. Blur and shadows cause most mistakes. Take the photo again in daylight, hold the phone parallel to the page and tap to focus.
• Boost contrast. It turns the image gray and stretches the contrast, which helps with photos, yellowed paper and light text on a dark background.
• Enlarge small images. Letters only a few pixels high are hard to read. This option scales small pictures up to about 1,500 pixels on the long side first.
• Straighten and crop. Turn a sideways photo with the rotate buttons, and select only the text you need: logos and patterns next to it produce stray characters.
• The right language. German umlauts, French accents and the Spanish ñ come out right only when that language is selected.

For a scanned image to text job, scan in grayscale at 300 dpi if your scanner lets you choose; that gives the engine enough detail. If nothing at all is found, the text is probably too small, too blurry or handwritten.

Free image to text with AI, without uploading

The recognition engine is Tesseract, an open-source OCR engine (optical character recognition) whose current versions read text with a neural network. Here it runs as WebAssembly inside your browser, and the picture is never sent to a server: not to LoschaZ and not to an AI service. That matters for ID cards, contracts, bank letters and medical forms.

Because nothing is processed on our side, there is no account, no daily limit and no watermark. The only downloads are the engine and the language file, both from this website. The language file is saved in your browser so it does not download again; clearing this site’s data removes it.

Google Lens or iPhone Live Text versus a browser tool

If the picture is already on your phone, the phone may be enough: iPhone and iPad can select text in photos with Live Text, and Google Lens does the same on Android. Both are quick for a phone number or a single paragraph.

A browser tool helps when you work on a computer, have several images or want the text as a file. Here you can read a stack of screenshots in one go, join the broken lines, correct the text in place and save it as .txt, without installing anything.

What happens to tables, columns and line breaks?

The result is plain text. Line breaks follow the lines in the picture, paragraphs are separated by an empty line, and bold, sizes and colors are not kept.

Tables become lines of text: depending on the spacing, a row comes out as one line or the columns are read one after another. Two-column pages are mostly read column by column, but a busy flyer can get mixed up; select one block at a time to keep the order.

Join broken lines turns the line breaks inside a paragraph into spaces and mends words split by a hyphen at the end of a line, so the text pastes cleanly into an email or a document. Afterwards, check words that really have a hyphen, such as well-known. The line above the result shows how many words were found and how sure the engine was on average; even when it is sure, check names, numbers and amounts.

More AI Text Scanner pages

• Picture to text: photos and the camera
• Screenshot to text: paste and copy