FAIND app · free tool
Image to text: copy the text from a photo or PDF
Free. Runs in your browser. Your images and PDFs never leave your device. Check it yourself ↓
Drop in a photo, a screenshot or a scanned PDF and get the text back, ready to copy or save. It runs in your browser, so the file is never uploaded. No account, no watermark, no limits. It reads printed text well and handwriting badly.
- No account
- No watermark
- No ads
- No page quota
OCR makes mistakes. Proofread the result, especially numbers, names and anything you will publish or sign.
Smart with PDFs, tidy with line breaks
A PDF that already contains real text is copied exactly, with no reading errors, and only scanned pages go through OCR. Once you have the text, "Join lines into paragraphs" removes the line breaks that come from the original layout, so it pastes cleanly into a document. You can also paste a screenshot straight from the clipboard and read several images in one go.
How it works
The page uses Tesseract, a long-established open-source OCR engine (Apache 2.0 licence) compiled to run inside the browser, and PDF.js from Mozilla to open PDFs. A photo is read directly. A PDF is checked page by page: if the page already holds real text, that text is copied as it is; if it is a scan, the page is drawn as an image and read like a photo.
Everything happens on your device. The engine is served by FAIND, and the first time you pick a language your browser downloads that language file (about 2 to 5 MB) from the jsDelivr CDN and keeps it. Your image or PDF is never sent anywhere. See the privacy notes.
Check it yourself: nothing is uploaded
You do not have to take our word for it. Any browser can show you where the data goes.
- Open your browser's developer tools (F12, or Cmd+Option+I on a Mac) and go to the Network tab.
- Choose a file and run the tool.
- Read the list of requests. You will see the reading engine loading and, the first time you use a language, its language file. After that your browser keeps them. No request carries your file: there is no upload (no large POST or PUT). The only other traffic is a cookie-free visit counter that sends the page address, never your file.
What people use it for
Copying text from a screenshot or a photo
A slide, a sign, a page of a book, a message you cannot select. Paste the screenshot, read it, copy the result.
Making a scanned PDF searchable or editable
Pull the text out of an old scan, a contract or a form, then paste it into a document. You get plain text, not a rebuilt layout.
Reading text in another language
Extract it here, then run it through a translator. Pair it with FAIND Audio to Text, which also translates.
What to expect
| Best results | Straight, sharp, well-lit printed text, dark on light, at 200 to 300 DPI. Screenshots are ideal. |
|---|---|
| Weak spots | Handwriting, blur, low light, curved pages, very small text, text over pictures, decorative fonts, rotated pages. |
| Layout | Plain text in reading order. Columns can run together and tables become lines. |
| Speed | A few seconds for a photo; roughly 5 to 15 seconds per scanned PDF page on a laptop, slower on a phone. |
| Limits | 25 MB per image, 50 MB and 50 pages per PDF. Keep the tab open while it works. |
| Languages | 19 languages and scripts, one at a time. The language is not detected for you. |
Questions people ask
How do I copy text from an image?
Drop the photo or screenshot into the box at the top of this page, pick the language of the text, and press Read the text. After a few seconds the text appears in a box you can edit, copy or save as a .txt file. You can also paste a screenshot straight from the clipboard. Nothing is uploaded: your own device does the reading.
Can I get the text out of a scanned PDF?
Yes. Each page is turned into an image and read the same way as a photo. If a PDF already contains real text (most PDFs made from Word or a website do), the page uses that text directly, which is faster and exact. Scanned PDFs are read at up to 50 pages per file; a long scan takes a few minutes.
Is my image or PDF uploaded anywhere?
No. The file is read by code running in your browser and is never sent to FAIND or to any other server. The page does download the recognition engine from FAIND and, the first time you use a language, a language file (about 2 to 5 MB) from the jsDelivr CDN. Your browser keeps it, so the next time it is much quicker.
Is it really free?
Yes. No account, no watermark, no limit on how many images you read. Your own device does the work, so it costs FAIND nothing to run.
How accurate is it?
On a clear, straight, high-contrast image of printed text, it is usually right on nearly every word. Accuracy drops on blurry photos, low light, curved pages, small text, busy backgrounds and unusual fonts. The page shows a confidence score so you know when to double-check. Always proofread before you rely on the text, especially numbers.
Can it read handwriting?
Not reliably. The engine is built for printed text. Neat block capitals sometimes come out well, but joined-up handwriting usually does not. If handwriting is what you need, test with a sample first and expect to retype a lot.
Which languages does it read?
English, Italian, French, German, Spanish, Portuguese, Dutch, Polish, Romanian, Turkish, Russian, Ukrainian, Greek, Arabic, Hindi, Japanese, Korean and Chinese (simplified and traditional). Pick the language of the text, because it is not detected for you. A page in a mix of two languages comes out better if you select the main one.
Why is the text full of strange characters?
Usually one of three things: the wrong language is selected, the image is too small or blurry, or the text is rotated or on a curved page. Choose the correct language, use a larger or sharper image, rotate it upright first, and crop out anything that is not text. A scan at 200 to 300 DPI works best.
Does it keep the layout, columns and tables?
No. You get the text in reading order, line by line. Two-column pages can be read across the columns, and tables come out as plain lines rather than cells. For a table, crop one column or one block at a time.
What file types and sizes work?
JPG, PNG, WebP, BMP, GIF and PDF, up to 25 MB per image and 50 MB per PDF. iPhone HEIC photos only open in browsers that can display them (Safari does); otherwise export them as JPG first, or take a screenshot of the photo.
Does it work on a phone?
Yes, in any recent browser on iPhone or Android. Reading is slower than on a computer, and a long PDF can be heavy for an older phone, so keep the tab open and the screen on while it works.
Can I read text from a photo of a screen, a sign or a menu?
You can try. It works best when the camera is square on to the text and the light is even. Glare, angles and decorative lettering are where it fails most. A screenshot is always cleaner than a photo of a screen.
How can I check that my image or PDF is not uploaded?
Open the developer tools of your browser (F12), go to the Network tab and read a file. You will see the engine and a language file being downloaded the first time, and no request that sends your image or PDF anywhere. The steps are in the "Check it yourself" section above.
Is it safe for contracts, ID documents or bank statements?
The file stays on your device and is not handed to any server, which is the main reason to prefer a tool like this for private documents. The extracted text also lives on your device, so protect it like the original and close the tab when you finish. OCR makes mistakes, so check every number before you rely on it.
How do I remove the line breaks from text copied from an image or PDF?
Tick "Join lines into paragraphs" under the result. Lines inside a paragraph are joined into one, blank lines between paragraphs are kept, and a word split by a hyphen at the end of a line is put back together. A compound word such as "well-known" that happened to break at the hyphen can lose its hyphen, so skim the result.
Can I turn a scanned PDF into a searchable PDF?
No. You get the text only, not a new PDF with a hidden text layer. Copy the text into a document, or save it as a .txt file, and keep the original scan as the reference.
Can I read several images at once?
Yes. Choose or drop several images together and the text of each one is added to the same result, with the file name above each part.
Read today's AI news Turn audio into text Find the BPM and key of a song

