Free Image to Text (OCR)
Pull editable text out of photos, screenshots and scanned documents — entirely in your browser. No signup, no upload, no page limit. Works in over 100 languages.
Language(s)
What are you scanning?
Advanced cleanup
Preparing…
No text detected. Try a sharper or higher-contrast image, crop to the text with “Select region”, or pick a document mode that matches the source.
Queue
Your file never leaves your computer
This tool runs entirely in your browser. There is no upload endpoint on this page — the PDF library does all of the work locally on your device, the result file is generated in memory, and the download starts from your own machine. Open your browser's network panel and verify: nothing is sent.
How it works
Three steps, no surprises
Add your image
Drop a file, paste, snap a photo or fetch a URL. Images and PDFs both work — nothing is uploaded.
Pick language & type
Choose one or more languages and the document type. The cleanup pipeline tunes itself to match.
Extract, edit, export
Fix any stray characters in the editor, then copy or download as TXT, Markdown, DOCX, PDF, searchable PDF or JSON.
FAQ
Image to Text (OCR) questions, answered
Is there a watermark, signup or page limit?
None of the above. Add an image, extract the text, copy or download it. There is no daily cap and no limit on the number of images.
Do my images get uploaded?
No. Recognition runs entirely in your browser via WebAssembly. The only network traffic is a one-time download of the OCR engine and the language model you pick, both cached locally for next time.
How accurate is it?
On clear printed text it is very accurate. Accuracy drops with low resolution, heavy skew, poor contrast, unusual fonts or background noise — the document modes and the cleanup options exist to help with exactly those cases.
Can it read handwriting?
Best-effort only. The engine is trained mainly on printed text, so neat, evenly-spaced handwriting may come through but cursive and messy notes often will not. Printed text is far more reliable.
How many languages are supported?
Over 100, including right-to-left scripts (Arabic, Hebrew) and CJK (Chinese, Japanese, Korean). You can tick more than one language at once for bilingual documents.
How big is the language download?
Smaller than most people expect, because the tool uses the compact LSTM models. English is about 2.8 MB; most languages are 1–3 MB and the largest are around 5 MB. Each model downloads once, then loads from your browser cache.
What is a searchable PDF?
A PDF that looks like your original scan but has an invisible, selectable text layer on top — so you can search and copy from it. This tool can export one directly.
Which document mode should I pick?
Start with Auto. Switch to Document for printed pages, Receipt for till receipts, Screenshot for screen captures, Table for rows and columns, Business Card for scattered fields, or Handwriting for handwritten notes.
Can I process several images at once?
Yes. Add multiple files and they queue up; extract them one after another and download everything together as a ZIP.
Why does the first extraction take longer?
The first run downloads the OCR engine and your language model. After that they are cached in your browser, so subsequent extractions start almost immediately.
Other free tools
Every one runs in your browser. No uploads, no signup.
Background Remover
AI removes the background from any photo in your browser. No watermark, no upload, no resolution cap.
Open →JPG to PNG
Convert JPGs to PNG with transparency support. Batch, drag-and-drop, no upload.
Open →PNG to JPG
Flatten PNGs to JPG for smaller file sizes. Quality slider, batch mode, in-browser.
Open →HEIC to JPG
Convert iPhone HEIC photos to JPG. Powered by libheif WebAssembly, runs in your browser.
Open →Need help with your website's SEO?
Book a free consultation with our London team.
Book a Free Consultation