Image to Text Converter

Extract text from images right in your browser. Recognition runs entirely on your device, so your image is never uploaded.

Runs 100% on your deviceReads Chinese, English, Japanese + 40 moreFree, no sign-up

Upload an image

Drag and drop, paste a screenshot (Ctrl+V), or click to browse

Supports JPG, PNG, GIF, WebP (max 10MB)

No image handy? Try a real receipt:

Bytes of your image sent to a server

0

Bytes of your image sent to a server

Languages read by a single model

50

Languages read by a single model

Model downloaded once, then cached

30MB

Model downloaded once, then cached

Export formats: TXT, Markdown, JSON

3

Export formats: TXT, Markdown, JSON

Coverage

Which languages can it read?

The built-in model recognizes 50 languages with a single set of weights: Simplified and Traditional Chinese, English, Japanese, and more than 40 Latin-alphabet languages including French, German, Spanish, Italian, Portuguese, Dutch, Polish, Turkish and Vietnamese.

Not supported yet

Korean, Arabic, Cyrillic and other non-Latin scripts are not recognized by the local model. For those languages, our server-side tool reads and translates the image instead: AI Image Translation

How it works

From image to editable text in three steps

1

Add your image

Drag and drop a file, paste a screenshot with Ctrl+V (Cmd+V on Mac), or click to browse. JPG, PNG, GIF and WebP up to 10MB.

The tool with a comparison table screenshot loaded and the Extract Text button ready
2

Text is recognized locally

The OCR model runs on your own machine. The first run downloads it once, about 30MB; after that extraction starts the moment you press the button, and it keeps working with no internet connection.

The tool recognizing text on-device, showing the progress bar and the one-time model download notice
3

Copy or download

Copy the result to your clipboard, or download it as TXT, Markdown or JSON. Turn on Preserve layout to keep columns aligned the way they sit in the image.

The extracted text with word, character, line, confidence and time counts, plus TXT, Markdown and JSON download buttons

Formatting

Keep the columns where they were

Plain extraction reads an image line by line, which scrambles anything laid out in columns. Turn on Preserve layout and every line is padded back to the character column it occupied in the image.

OffColumns collapse into one another
ALEXANDER'S HANOVER ST
THE"SUPER" MARKET
11/18-12:21PM STORE 10
CUST 94 REG 8 OPR 171
GROCERY :60
HOOD HOODSIE CUP 1.99
HOOD HOODSIE CUP 1.99
HF ORANGE JC 64Z 1.29
BC WHITE CRKE 18 .73 一
BC WHITE CRKE 18 .73
SLICE D/MAN OR2L .89
S/K NAPKINS 140
OnRows and columns line back up
  ALEXANDER'S HANOVER ST
     THE"SUPER" MARKET
  11/18-12:21PM STORE  10
  CUST 94 REG 8 OPR    171
     GROCERY        :60
HOOD HOODSIE CUP   1.99
HOOD HOODSIE CUP   1.99
HF ORANGE JC 64Z   1.29
BC WHITE CRKE 18    .73 一
 BC WHITE CRKE 18   .73
 SLICE D/MAN OR2L   .89
 S/K NAPKINS 140

Both blocks are the real output of running the English receipt sample above through this tool. The toggle affects the on-screen result and the TXT download.

Privacy

Why local OCR is different

Most online OCR tools upload your image to a server for processing. This tool brings the model to your browser instead, so the image itself never travels.

This tool
Where your image goes
Stays in your browser
Internet connection
Only for the first model download
Usage limits
None
Account
Not needed
Typical online OCR
Where your image goes
Uploaded to a server
Internet connection
Needed for every image
Usage limits
Daily quotas or paid credits
Account
Often required

You can check this yourself: run one extraction, disconnect from the internet, and the tool keeps working.

Use cases

What people use it for

Screenshots and chats

Paste a screenshot straight from your clipboard and pull the text out of error messages, chat threads or presentation slides. No saving a file first.

Receipts and documents

Photograph a receipt, invoice or letter and turn it into searchable, editable text. Sensitive papers never leave your device.

Book pages and printouts

Digitize pages from books, handouts and printed notes without retyping a word. Extraction takes a second or two per page.

Tables inside images

This page outputs plain text. If your image contains a spreadsheet-style table, use the dedicated tool instead: image to table converter

Next step

Need the text in another language?

Extraction gives you the original text. These tools pick up where OCR leaves off.

AI Image Translation

Upload an image and get it back translated with the layout preserved. Supports Korean and many other languages that local OCR cannot read.

Translate an image

OCR Translation

For scanned PDFs and photographed documents that need text recognition and translation in one pass.

Translate a scanned document

Questions

Frequently asked questions

How the tool works and what it can read.

Is my image uploaded to a server?

No. The OCR model runs inside your browser using WebAssembly. Your image is processed on your device and never leaves it. The only download is the model itself, which your browser fetches once and caches.

How do I copy text from a screenshot?

Take the screenshot, then press Ctrl+V (Cmd+V on Mac) anywhere on this page. The image loads instantly and you can extract and copy the text without saving a file first.

Is it really free? Are there any limits?

Yes. There are no daily quotas, no watermarks, no sign-up and no paid tier for this tool. Extract as many images as you like.

Which languages are supported?

Simplified and Traditional Chinese, English, Japanese, and over 40 Latin-alphabet languages such as French, German, Spanish, Italian, Portuguese, Dutch, Polish, Turkish and Vietnamese. Korean, Arabic and Cyrillic scripts are not supported yet.

Can it read handwriting?

It is built for printed text. Clear, neat handwriting sometimes works, but accuracy drops quickly, so treat handwriting results as a draft that needs checking.

What image formats and sizes work?

JPG, PNG, GIF and WebP files up to 10MB. For best results use a sharp image where the text is not too small and the contrast between text and background is high.

How do I export the result?

Copy it to the clipboard, or download it as plain text, Markdown or JSON. Enable Preserve layout first if you want columns to stay aligned, which helps with receipts and printed tables. The JSON export includes the position and confidence score of every detected line.

Can I extract a table from an image?

This page outputs plain text. For spreadsheet-style images, use our image to table converter, which rebuilds rows and columns and exports them to Excel.

Why does the first run take longer?

Your browser downloads the recognition model (about 30MB) on first use. It is cached afterwards, so later extractions start immediately and even work offline.

Embed Image to Text Converter on Your Website

Add our free OCR tool to your website with a simple iframe embed. Works offline, protects privacy, and supports multiple languages.

e.g., 100%, 800px

Minimum 600px recommended

<iframe
  src="https://doc2lang.com/embed/image-to-text"
  width="100%"
  height="800px"
  loading="lazy"
  allow="clipboard-read; clipboard-write"
  style="border:1px solid #d1d5db;border-radius:8px"
  title="Doc2Lang image to text converter"
></iframe>

Usage Tips

  • • Copy the code above and paste it into your website's HTML
  • • The widget runs completely in the user's browser with full privacy
  • • Customize width, height, and default language to match your needs