To extract text from an image for free, open Extract text on AIMRAN Tools, choose the language of the text, drag in the photo or screenshot (or paste it with Ctrl+V) and click “Extract text”. A few seconds later the text is ready to copy or download as a .txt file. Recognition (OCR) runs in your browser, so the image is never uploaded to a server. If what you have is a scanned PDF, use OCR for scanned PDFs, which also gives you back the same PDF with searchable text.
What OCR is and when you need it
A photo of a document, a screenshot or a scan is an image: to your computer there are no letters, only pixels. That is why you cannot select the text, search for it or paste it into Word. OCR (optical character recognition) analyses those pixels, identifies the letters and turns them into real text.
Typical cases: copying the text from a screenshot into an email, typing up photocopied notes without retyping them, noting down the figures on a receipt for an expense sheet, or recovering a scanned letter.
How to get text from an image online, step by step
- Go to tools.aimran.es/en/imagen/ocr on your computer or phone.
- Drag the image onto the “Drag photos or screenshots with text” area, click to pick it, or paste a screenshot with Ctrl+V. You can add several at once, and more later with “Add more images”.
- In the “Text language” panel, choose “Spanish”, “English” or “Both”.
- Click “Extract text”. The first time, the language data is downloaded (about 2.1 MB for Spanish, 2.9 MB for English and 5 MB for both); after that it is cached and the tool starts instantly.
- Check the confidence label for each image, then copy the result with “Copy” or save it with “Download” as
texto-extraido.txt.
If you add several images, all the text comes out together, with each file name as a separator, so you know where each part came from.
It accepts JPG, PNG, WEBP, AVIF, BMP, GIF and HEIC (iPhone photos), up to 30 MB each.
How to read the confidence score
Each image shows a percentage that tells you how sure the engine is about what it read:
| Confidence | Colour | What to do |
|---|---|---|
| 80% or more | Green | The text is usually right; give it a quick read. |
| 55% to 79% | Orange | Some letters will be mixed up. Check numbers, accents and names. |
| Below 55% | Red | Better to try again with a sharper, cropped or straightened image. |
If you see “No text was found”, the image is too small, blurry, or you picked the wrong language.
Extracting text from a scanned PDF
First check whether your PDF needs it: try to select a word with the mouse or find it with Ctrl+F. If that works, the PDF already contains text and Extract text copies all of it without OCR. If it doesn’t, each page is a picture and needs to be recognised:
- Open OCR for scanned PDFs and drag in the PDF (up to 200 MB).
- Choose the “Document language”.
- Leave “Create searchable PDF” on if, besides the text, you want a PDF that looks exactly like the scan but can be searched and copied.
- Click “Recognise text”. The button tells you how many pages it will process.
- Copy or download the “Recognised text” (with “Separate by page” on you will see where each page starts) and, if you asked for it, download the searchable PDF, whose name ends in
-buscable.pdf.
Each page is rendered as an image at about 300 DPI before it is read, which is the resolution Tesseract recommends. The text layer is placed invisibly on top of the original page, without touching the image.
If you only need a few pages of a long document, pull them out first with Extract pages: the OCR will finish much sooner.
Tips for good OCR results
The Tesseract documentation (Tesseract is the open-source engine behind these tools) stresses the same things you notice in practice:
- Enough resolution. It works best with images of at least 300 DPI. When you take a photo with your phone, get close so the text fills most of the frame.
- Straight text. Skewed lines make results much worse. Turn the photo with Rotate image if it is sideways.
- Crop, but not too tight. Remove backgrounds, fingers and other areas with Crop, leaving a small margin around the text; text touching the edge also causes errors.
- Good light and contrast. Shadows and glare on the paper confuse the engine. Dark text on an even, light background is best.
- One language. Choosing only the document’s language gives better results than “Both”.
Common mistakes
- Handwriting. OCR works very well with printed or on-screen text, but handwriting is not recognised reliably.
- Wrong language or unsupported language. The tool recognises Spanish and English. Text in other languages, such as French or German, usually comes out with errors in accented letters.
- Tables and columns. The text comes out in reading order, but the structure of a table is not kept. If you only need one column, crop it first.
- Password-protected PDF. If the PDF asks for a password when you open it, remove it first with Protect & unlock.
- Getting the text into Word. You don’t need a special converter: copy the result and paste it into Word, or open the downloaded .txt file.
How to copy text from a photo on iPhone
If it is a single photo on your iPhone, the system can do it with Live Text. According to Apple’s official help, open the photo in the Photos app, touch and hold a word, adjust the selection with the grab points and tap Copy (or Select All). You need an iPhone XS or XR or later with iOS 15 or later, and Live Text turned on in Settings > General > Language & Region.
For several images at once, multi-page PDFs, downloading the result as .txt or working on a computer, the browser tool is more convenient.
Conclusion
Getting the text out of an image doesn’t require installing anything or uploading your documents to an unknown website. With a sharp, straight photo, the right language and a glance at the confidence score, you have the text ready to copy in a minute. And if the document is a scanned PDF, the PDF OCR also gives you a searchable copy that looks just like the original.