Image to Text (OCR)

What is this image to text tool?

This image to text tool is an OCR (optical character recognition) tool that reads the text in photos, screenshots, and scanned documents and turns it into text you can copy and edit. No more retyping paper documents or copying on-screen text by hand.

It uses a text recognition model that reads both Japanese and English. Everything runs inside your browser, and your images are never uploaded to a server. There is no sign-up or login, and you can use it as often as you like for free.

How to extract text from an image (3 steps)

1. ChooseDrop the image onto the "Drop an image here, or click to choose" box, or click it to pick a file. You can also paste a screenshot with Ctrl+V (⌘+V on Mac).
2. See the resultText recognition starts automatically once the image loads. The first time, the text recognition model (about 30 MB) is downloaded, so it takes a little longer.
3. Fix and exportWhen the text appears, fix any mistakes in the "Recognized text (editable)" box, then use "Copy text" or "Save as .txt".

Supported images and languages

The tool reads images of horizontal printed text. Text that mixes Japanese and English is read as is.

Image formats

JPEG, PNG, WebP, BMP, GIF (one image at a time)

Languages

Japanese (hiragana, katakana, kanji) and English, plus numbers and symbols

Works well for

Photos of documents and business cards, scanned pages, screenshots of apps and web pages, and error messages

Tips for more accurate results

Make the text large and sharp

Small or blurry text leads to more mistakes. When shooting with a phone, move close so the text fills the frame and wait for it to focus.

Shoot straight on, without shadows

Slight tilt is corrected automatically, but documents shot at a steep angle or covered by shadows and glare are harder to read. Shoot straight on in good light.

Crop to the part you need

For images that mix photos, diagrams, and text, crop out just the part you want with Image Cropper first so unrelated text is not picked up.

What the browser version struggles with

This is a lightweight version that runs only in your browser. It reads printed horizontal text and on-screen text accurately, but it struggles with images like these:

Messy handwriting

Joined-up or hurried writing is often missed or misread. For neat handwritten notes, use Handwriting to Text.

Vertical Japanese text

Vertical text in books, newspapers, and greeting cards is either not recognized or comes out with many errors.

Complex tables or columns

The reading order of the extracted text may break.

Image to Text FAQ

Is this image to text converter free?

Yes. There is no sign-up, no login, and no usage limit. Nothing to install either: it runs in your browser.

Are my images uploaded to a server?

No. Text detection and recognition both run entirely inside your browser. The only network request is the one-time download of the text recognition model (about 30 MB). Neither your image nor the extracted text is sent anywhere.

Can it read handwriting or vertical Japanese text?

This tool is built for horizontal printed text, such as documents and on-screen text. For handwritten notes, use Handwriting to Text, which comes with tips for photographing handwriting. Vertical text is either not recognized or comes out with many errors.

Can I extract text from a scanned PDF?

This tool reads image files (JPEG, PNG, WebP, BMP, GIF) and cannot open PDFs directly. Export the pages of a scanned PDF as images first, then load them here.

Does it work on my phone?

Yes. Open the page in your phone's browser and pick a photo or screenshot. If an iPhone photo in HEIC format will not load, convert it to JPEG with HEIC to JPG/PNG/WebP Converter first.

What should I do if the result has mistakes?

The "Recognized text (editable)" box can be edited directly, so fix any wrong characters before copying or saving. If there are many mistakes, retake the photo so the text fills the frame, or crop the image to just the part you need.