Image to text (OCR)
Turn the text in a photo or scanned PDF into text you can edit and copy. Files are processed on this device and never uploaded.
Add one image or PDF that contains text.
or drop a file here
Recognition is automatic and can get letters wrong, so check anything important against the original.
How to use it
- Click "Choose file" or drop one image or PDF that contains text.
- Under "Document language", tick every language used in the document, for example English and Spanish for a mixed page.
- For a PDF, choose "All pages" or a "Page range".
- Click "Extract text", fix any wrong letters in the result box, then "Copy" or "Save as TXT".
What it does
Photos and scanned PDFs
Reads JPG, PNG, WebP and HEIC photos and screenshots, and goes through scanned PDFs page by page.
Seven languages
Choose English, Korean, Japanese, German, French, Spanish or Portuguese, or several at once for mixed documents.
Edit before you copy
The text appears in an editable box, so you can correct misread letters before copying it or saving a TXT file.
No upload
The recognition engine runs inside your browser, so contracts, receipts and IDs stay on your device.
Pick the pages
For long PDFs, read only the pages you need, such as 1-3, 5. Each page is labelled in the result.
Good to know
OCR stands for optical character recognition: software that reads letters in a picture and turns them into real text. It saves retyping when the words are stuck in an image, such as a photo of a printed page, a screenshot of a slide or a scanned contract.
How well it works depends mostly on the picture. Fill the frame with the text, keep the page flat and avoid shadows and blur. Handwriting, very small print and complex tables come out less accurately, and table layouts are not kept. The first run downloads the recognition engine and the language data from this site (about 1 to 3 MB per language); after that your browser reuses them.
Questions and answers
Is this image to text converter free?
Yes. There's no sign-up, and your images are read inside your browser rather than uploaded to a server.
Can it read scanned PDFs?
Yes. Each page of the PDF is drawn as an image and then read. For long files, use "Page range" to read only the pages you need, which is much faster.
What if the document mixes two languages?
Tick both languages under "Document language". Each extra language slows recognition down, so only pick the ones that actually appear.
Why are some letters wrong?
Automatic recognition makes mistakes, especially on blurry or tilted photos. Correct them in the result box, or retake the picture closer and in better light. Always double-check names and numbers against the original.
Does it work with handwriting?
Neat handwriting is sometimes partly read, but accuracy is much lower than for printed text, because the engine is trained on printed letters.
Why does the first run take longer?
The recognition engine and the language data are downloaded from this site the first time. Only program data is downloaded; your file never leaves your device.