Free Image to Text Converter
nothing stored

What this image to text converter does

This tool reads the printed words in a picture and gives them back as plain text you can copy, search and edit. The technique is called OCR, short for optical character recognition.

It works on the kinds of images people actually have:

- Screenshots of chats, emails, web pages and apps that do not let you select text. - Photos of documents, book pages, menus, labels and signs. - Scans of letters, forms and receipts.

It reads 10 languages: English, Spanish, French, German, Portuguese, Italian, Hindi, Chinese (Simplified), Japanese and Arabic. You choose one per run.

The result shows the text with a Copy button, plus the number of words and lines and a confidence score: how sure the recognizer was about the words it kept. You can also download the text as a .txt file. Everything happens in your browser.

How to extract text from an image

1. Add your image. JPG, PNG, WebP and iPhone HEIC photos all work. Crop away anything that is not text if you can; less clutter means fewer mistakes. 2. Choose the language of the text. Reading French text with English selected loses the accents and misreads some words. 3. Choose a layout. Automatic handles pages with paragraphs and columns. Single block suits one paragraph, a receipt or a cropped screenshot. Scattered text finds words dotted around an image, like a sign in a street photo or labels on an app screen. 4. Press Extract text. The first time you use a language, your browser downloads its reading data (under 3 MB) and keeps it, so later runs start right away. 5. Copy the text or download it as a .txt file.

If a line comes out wrong, try the other layouts before anything else. It is the setting that changes results the most.

Getting accurate results

The recognizer is only as good as the picture it gets. Before reading, the tool already does three things for you: it enlarges small text, stretches faded contrast, and smooths grain in dark or noisy photos. The rest is up to the photo.

Get close and fill the frame. Letters should be clearly readable to you when you zoom in. Text that is a few pixels tall in a wide photo is hard for any reader.

Shoot straight on. A page photographed at a steep angle bends the lines. Hold the phone parallel to the page.

Avoid glare and shadows. A reflection across a glossy page wipes out the words under it.

Keep it upright. Slightly tilted text is fine, but a photo lying on its side should be rotated first, for example with the free rotate tool.

Check the confidence score. Above 90% usually means clean text. Well below that, proofread names, numbers and prices before you rely on them.

What it can and cannot read

It reads well: printed and on-screen text in the 10 supported languages, in ordinary fonts, dark on light or light on dark. Screenshots are the easiest case because the letters are sharp.

It struggles with:

- Handwriting. It is built for print. Neat block capitals sometimes work; joined-up writing mostly does not. - Decorative fonts such as heavy scripts, outlined letters or text over a busy photo. - Tables and forms. You get the words, but not the columns and cells. Paste into a spreadsheet and tidy by hand. - Mixed languages. One language is read per run. A Japanese menu with English prices is best read twice, once in each language. - Very small or blurry text, where the letters are not clear to a person either.

When nothing readable is found, the tool says so instead of returning a page of guesses. It also drops fragments it was unsure about, such as texture in a photo that looks a little like letters.

Reading Text From Documents Without Uploading Them

The images people want text from are often the ones they would least like to send anywhere: bank letters, ID documents, medical forms, private chats, contracts and receipts with card details.

This converter reads the text in your browser, on your own device. Your image is not uploaded and nothing is saved on our servers. The only thing that is downloaded is the reading data for the language you chose, the same small file for everyone, which your browser keeps so it is not fetched again.

Many free OCR websites work the other way round: you upload the image, a server reads it, and the text comes back. That can be quick, but a copy of your document has left your hands.

Processing is unlimited. Your first few downloads need no account; after that, a free sign-in keeps them going. Copying the text never needs one, and signing in never uploads your image.

Copying text vs changing text with AI

The free Image to Text converter takes words out of a picture. You get plain text to paste into a document, email, translator or spreadsheet. The picture itself is not changed.

Sometimes the goal is the opposite: the picture is right, but the words in it are wrong. A shop sign with an old phone number, a product label in the wrong language, a menu photo with last year's prices, a poster with a typo.

For that, Imagera's paid AI image editor changes the picture. You describe the edit, for example "change the sign to say Open Sundays", and it redraws that part of the image to match the lighting and style around it. It runs in the cloud and uses credits.

A simple workflow uses both: extract the text here, fix or translate it, then ask the editor to put the corrected words back into the picture.