Sharper image, better accuracy
Use straight, well-lit images with printed text. Handwriting and blurry photos can produce errors.
Turn images into editable text
Extract, review and export the text from images and PDFs. All in one place.
Your document
PNG · JPG · WEBP · BMP PDF
Click or drag your file hereOne file at a time · up to 50 MB
On the first extraction, the recognition engine and the languages are downloaded. The file’s content stays on your device.
Extracted text
0 characters · 0 words
PDFs with many pages
Select a range of up to 100 pages. PDFs with selectable text are read directly, which speeds up the extraction.
Spreadsheets to review
Excel separates columns by tabs or two spaces. The structure is approximate: review complex tables. Word exports editable text, without reproducing the original layout.
About OCR
For copying the text from photos, screenshots and scanned PDFs without typing anything: receipts, book pages, old documents.
What this OCR does
It works in your browser by default and only uses AI if you ask for it. The features that stand out:
Local by default, AI only if you turn it on
Recognition runs in the browser, without sending the image. For handwriting and difficult photos, the “Try with AI” box sends the image to an AI service, and the page warns you first.
Exports to Word and Excel, besides TXT
Copy the text or download it as TXT, DOCX or XLSX. Excel separates columns by tabs or two spaces (the structure is approximate) and Word carries the editable text, without reproducing the layout.
PDFs of up to 100 pages
Choose a page range. PDFs with selectable text are read directly, which speeds up the extraction.
Keeps spaces to separate columns
An option keeps the original’s spaces, useful for tables, receipts and lists in columns.
Six language options
Portuguese, English, Spanish, French and German, plus a combination of Portuguese, English and Spanish.
Editable text before exporting
Correct numbers and names in the text box itself before copying or downloading.
A practical example
You took a photo of a receipt to enter its data into a spreadsheet. You upload the image, choose the language and let the recognition run. The text appears; you fix a number that came out swapped and export to Excel or copy it into the spreadsheet.
How to use it
- Upload the image or the PDF.
- Choose the text’s language and the mode (local or optional AI).
- Review the recognized text and correct whatever is needed.
- Export as TXT, Word or Excel.
Limits and cautions
- Shaky, dark or tilted photos reduce accuracy; prefer sharp, straight images.
- Handwriting works better with the optional AI mode.
A ferramenta em uma imagem
Os pontos de clique estão numerados. Toque na imagem para vê-la em tamanho original. As demais telas, passo a passo, estão no guia “Como funciona”.
- Point 1: Upload the image or PDF: click the area or drag the file (PNG, JPG, WEBP, BMP or PDF up to 50 MB).
- Point 2: Choose the language of the text: it changes the result a lot, especially with accents.
- Point 3: Extract text: in local mode, recognition runs in your browser.
- Point 4: After reviewing the text, copy it or export it as TXT, DOCX or XLSX.
When to use it
- Copying the text from a photo, screenshot or scanned document instead of typing it.
- Digitizing receipts, bills and proofs of payment into spreadsheets.
- Making the text of a scanned PDF searchable.
Tips
- Photograph in good light, straight on and without shaking; crop the useless edges.
- Choose the right language: it changes the result a lot, especially with accents.
- Review numbers and proper names; those are the places where recognition errs the most.
What happens in your browser
OCR (optical character recognition) is the technique that turns the image of a text into real text. In local mode, recognition runs in your browser, with an engine that identifies the letters and words in the image and uses the chosen language to get more right.
Accuracy depends on the image: sharp, straight text with good contrast and good resolution gives much better results. Shaky, dark or tilted photos reduce the hits.
There is also an optional AI mode, which helps with handwriting and hard cases. When you turn it on, the image is sent to an outside service; that is why the page warns you first. Without it, nothing leaves your browser.
Read the full guide: real-world use cases, tips and limits →
Privacy: It works locally by default. Only if you turn on the AI feature is the image or PDF page being analyzed sent to Groq (vision model). The standard mode runs 100% locally (Tesseract.js, in the browser). The image only leaves the device if you tick the “Try with AI” box — meant for handwriting and difficult photos. Learn more.
Perguntas frequentes
- Is my document sent to a server?
- In local mode, no: recognition runs in your browser. Only the AI mode, which you turn on explicitly, sends the image for processing.
- Why did the text come out with errors?
- OCR depends on the quality of the image. Increase the resolution, improve the lighting or try the AI mode.
- Does it work with handwriting?
- It works better with the optional AI mode; local mode is better suited to printed text.
- Which formats can I export to?
- TXT, Word and Excel.
- Does it work with a scanned PDF?
- Yes. A scanned PDF is made of images, and OCR turns them into text.