What this tool does
HANECopy is a free OCR tool that pulls text off your computer screen — just draw a box around it. No install or sign-up required. All text recognition takes place inside your browser, and your screen is never sent to or stored on a server.
Typical OCR services convert an image or PDF file that you upload. This tool works the other way around: it reads whatever is on your screen right now. That means it can extract text that isn't saved as a file at all — such as business app screens or the wording of an error dialog.
-
🔲Just draw a box
Place the blue box over the text and click the button — what's inside becomes text that you can copy with one click.
-
🌐Choose a reading mode
Five modes: Japanese + English, Japanese only, English only, error messages/logs, and Japanese phone numbers (digits only). Recognition covers Japanese and English (Latin letters and digits) only — other languages are not supported. Picking a narrower mode improves accuracy.
-
🔁Auto-watch
Reads the same area repeatedly at a set interval and alerts you only when the content changes — handy for waiting until a value on screen updates.
-
💧Color code picker
Click anywhere on the shared screen to get that color's code in both HEX and RGB.
The recognition engine is a browser port of the open-source Tesseract OCR. The language data (Japanese and English) is hosted on this site too, so no external service is ever contacted.
How to use it (3 steps)
-
Choose a screenClick “Choose a screen to read” and your browser will ask you to share your screen. Select the window you want to read, or your entire screen.
-
Box the textIn the preview, move the blue box over the text you want. Drag inside the box to move it and drag the bottom-right corner to resize it.
-
Read and copyClick “Read & copy now” and the text inside the box is converted and copied. Turn on Auto-watch if you want it to keep reading at the interval you choose.
When it comes in handy
Copying text you can't select on screen
If it's on your screen, it can be read — regardless of file format or application.
Some business systems and applications don't let you select the text they display. The same goes for error dialog messages, text embedded in viewers, and tables pasted as images. The only way to move that text elsewhere is to retype it, and the longer the data, the more likely typos become.
With this tool, anything visible on screen can be read. Because it doesn't depend on file formats or application types, there is no software it can't work with.
Copying phone numbers from call-center (CTI) or business software
Phone number mode limits recognition to digits and hyphens, greatly reducing misreads.
Call-center (CTI) screens that show the caller's number often don't let you copy it. Setting the mode to “Japanese phone number” limits recognition to digits, hyphens and similar characters, which greatly reduces misreads. Note that this mode only accepts Japanese-format numbers (10 or 11 digits starting with 0, or +81). You can choose whether to keep hyphens to match the format your destination system expects.
With Auto-watch on, the number is read automatically whenever it changes, so you don't have to do anything for each incoming call.
Copying tables and lists from the screen
Box several lines at once and the line breaks are preserved. If columns get misaligned, read one column at a time.
Box a multi-line list in one go and it's extracted as text with the line breaks intact, ready to paste into a spreadsheet. If columns get misaligned, reading one column at a time makes it easier to tidy up.
To clean the data up as a CSV, paste it into our CSV editor, which keeps leading zeros intact.
OCR basics
What is OCR?
It's technology that lets a computer read the characters in an image and convert them into text you can edit and search.
OCR stands for Optical Character Recognition: technology that lets a computer read the characters in an image and turn them into editable, searchable text. It has long been used to digitize scanned paper documents.
In an image, characters are just a collection of dots. OCR analyzes their shapes and decides “this shape is the letter A.” Being readable to a human doesn't guarantee OCR can read it — results depend on the text size and its contrast with the background.
Tips for better accuracy
Zoom in so the text is larger, shrink the box to just what you need, and pick a narrower mode. These three make a big difference.
A few simple tricks can greatly improve recognition accuracy.
- Display the text larger — zoom in on the source application before reading. The smaller the text, the harder it is to tell characters apart.
- Keep the box tight — stray decorations or lines inside the box cause misreads. Leave just a little margin around the text.
- Narrow the mode — choose “English only” for English text, or “Japanese phone number” for Japanese phone numbers. Fewer candidate characters means fewer mistakes.
- Ensure good contrast — black text on white is the most reliable. Text over a background image or in light gray is recognized less accurately.
Choosing a reading mode
Matching the mode to the content reduces the candidate characters and improves accuracy.
This tool has five reading modes, selected from the menu at the top. The mode changes the result, so pick the one that matches what you're reading. The tool recognizes Japanese and English only.
- Japanese + English — the default. Suited to screens that mix Japanese and English. With the most candidate characters, it takes a little longer.
- Japanese only — for screens with only Japanese text. Excluding Latin letters improves accuracy for kanji and kana.
- English only — for English text, model numbers, URLs, IDs and anything else made of Latin letters and digits. It can also pull just the English out of a screen that mixes in Japanese.
- Error messages / logs — for copying error dialogs and logs exactly as shown. Spaces between words are kept and symbols are left as read, so error codes and file paths stay intact.
- Japanese phone number — a special mode that limits recognition to digits, hyphens and the like. It returns a result only when it finds a Japanese-format number (10 or 11 digits starting with 0, or +81), so ordinary text never comes out as a string of random digits. Phone numbers in other countries' formats are not detected — use “English only” for those.
When different kinds of data sit side by side on one screen, reading them in separate passes and switching the mode each time gives more stable results. For example, where a Japanese name sits next to an alphanumeric ID, reading them separately is more accurate than doing both at once.
The Japanese modes also include data for vertical writing, so text can be extracted even from vertically typeset Japanese documents.
Characters OCR often confuses
Look-alike characters such as “1” and “l” or “0” and “O” — and in Japanese, pairs like ロ/口 and カ/力 — are easily mixed up.
Japanese uses thousands of different characters, which makes it harder to recognize than English. Look-alike characters are especially easy to confuse — typical examples are katakana ロ vs. kanji 口, katakana カ vs. kanji 力, and katakana エ vs. kanji 工.
Mixes of digits and letters need care too. “1” and lowercase “l”, or “0” and uppercase “O”, look almost identical — even people can't tell them apart without context. For strings with no meaning, like invoice or model numbers, we recommend checking the result after reading.
Small marks such as accents or Japanese voicing marks can be dropped, and thin fonts or tightly packed layouts also reduce accuracy. Always check the result with your own eyes before using it.
Can it read handwriting?
Neatly written block digits can be read to some extent, but cursive or messy writing won't reach a usable accuracy.
Handwriting is recognized far less accurately than printed text, because letter shapes vary from person to person and strokes connect inconsistently. Neatly written block digits can be read to some extent, but cursive or messy writing won't reach a usable accuracy.
For accurate handwriting recognition you'll need a paid AI-based OCR service. This tool is best at quickly turning printed text shown on screen into text.
How it compares with other OCR tools
OCR tools fall into three broad types. The biggest difference is whether they read a file or your screen.
| Installed OCR software | Online OCR (upload) | HANETOOL (screen reading) | |
|---|---|---|---|
| What it reads | Image/PDF files | Image/PDF files | Whatever is on screen |
| Text that isn't a file | Can't read | Can't read | Can read |
| Where data is processed | On your computer | Provider's servers | Only in your browser — nothing is sent |
| Setup | Install required | None | None — just open it in a browser |
| Usage/size limits | Depends on product | Often capped | None (free version) |
| Batch processing | Strong | Strong | One area at a time — move the box and repeat |
* This summarizes general tendencies of each approach and does not compare specific products. Terms vary widely between products and services, so check each one's terms before use.
Broadly, OCR tools are online services, software you install on your computer, or tools like this one that run entirely in your browser. Each has strengths and weaknesses, so choose based on the data you handle.
Compared with online OCR (no upload needed)
Recognition runs entirely in your browser, so nothing is uploaded. Your screen never leaves your computer.
Most online OCR services process your file by uploading it to the provider's servers. That's convenient, but it means your content passes through someone else's server. For contracts or documents containing customer data, company policy often forbids uploading to online services.
This tool is also opened online, but recognition happens only inside your browser, so nothing is uploaded. Your screen and data never leave your computer, and results aren't recorded anywhere. That's why it can be used even where online OCR is restricted.
The only network traffic is the first download of the language data, which simply fetches files hosted on this site and has nothing to do with what you read. Once loaded, recognition runs entirely in your browser, just as it would offline.
Free online OCR services often cap the number of conversions per day or the file size. This tool uses your own computer's processing power, so there are no such limits — read as much as you need without worrying about a free quota.
No installation required
Just open it in your browser. Because it uses screen capture, Chrome or Edge is recommended.
On work computers you often lack admin rights and can't add software. In many workplaces, installing OCR software requires a request and a long wait for approval.
This tool works just by opening it in a browser. Because it relies on screen capture, Chrome or Edge is recommended. Smartphone browsers don't support screen capture, so please use a computer.
Reading several spots in a row
The box keeps its position and size, so with similar layouts you can just nudge the box and keep extracting.
To pull several items from one screen, move the box and read again. The box keeps its position and size, so when similar layouts are lined up, you can just shift the box little by little and keep extracting.
To collect from several windows, switch the source with “Change screen.” Your mode setting is kept, so there's no need to set it again each time. Up to 20 extracted texts stay in the history, and one click copies them again — so even after reading many items, you can go back and pick the one you need.
For routine forms where you extract items from the same position every time, combining this with Auto-watch saves even more effort.
Reading image or PDF files
This tool can't open files directly. Use the text extraction feature of our PDF editor instead.
Since this tool reads what's displayed on screen, it can't load image or PDF files directly. You can open a file and read it on screen, but it isn't suited to processing long PDFs in bulk.
To extract text from a file, our PDF editor has a text extraction feature. It supports both PDFs and images and lets you extract text from a selected area. Use whichever suits the job.
About the color picker
Click “Pick color,” then click anywhere on the shared screen to see that pixel's color in HEX (#RRGGBB) and RGB. Click either one to copy it.
Useful for checking a website's color scheme, matching colors across documents, or reusing a logo's exact color. Unlike browser developer tools, it works on anything on your screen — a design app, a frame of a video, or a figure inside a PDF.
FAQ
Can I read text on my screen without installing OCR software?
Yes — HANECopy is free, with no sign-up, no install and no usage limit.
Yes — HANECopy is free. No sign-up or install is needed, and there's no limit on how many times you can use it. You draw a box around text shown on your screen and it becomes text. All recognition happens in your browser, so nothing has to be uploaded anywhere. If you're looking for OCR that converts a file instead, our PDF editor has a text extraction feature.
How is this different from text recognition in Google Drive or Google Lens?
The big difference is whether you have to upload your file to Google's servers.
In Google Drive, if you upload an image or PDF and open it as a Google Doc, the text is extracted for you. Google Lens can also read text in images. Both are free, but they require uploading your file to Google's servers. If you work with data that must not leave your organization, or want to read on-screen text that isn't a file at all, this tool — which processes everything in your browser — is a better fit.
When should I use this instead of the Windows 11 Snipping Tool?
Snipping Tool can recognize and copy text, but it has no reading modes or repeated watching.
The Windows 11 Snipping Tool can recognize and copy text in a snipped area, and since it's built in, it's convenient. However, it can't narrow down the recognition language or repeatedly watch the same area. If you want to read only the digits of a phone number, or wait for a value on screen to change, this tool is likely the better choice.
Does it work on smartphones?
This tool is for computers only. On a phone, Google Lens or iPhone's Live Text work well.
This tool relies on screen capture, which smartphone browsers don't support, so it's for computers only. On a phone, built-in features such as Google Lens or iPhone's Live Text are handy — just point the camera and select the text to copy. For text on a computer screen, use this tool.
How do I extract text from a PDF?
Our PDF editor has a text extraction feature that supports both PDFs and images.
Our PDF editor has a text extraction feature. Open a PDF and drag over an area to see its text. PDFs that contain text are processed in your browser, and OCR is used only for scanned (image) PDFs. Image files are supported too. If you just need one page quickly, you can also display the PDF on screen and box it with this tool.
How can I improve OCR accuracy?
Zoom in, tighten the box, and switch modes. Black text on white is the most reliable.
First, zoom in on the source so the text is displayed larger — the smaller the text, the harder it is to recognize. Next, shrink the box to just the part you want so no lines or decorations get in. If you know what kind of text you're reading, switching the mode to “English only”, “Japanese only” or “Phone number” reduces the candidate characters and cuts down on misreads. Black text on a white background gives the most stable results.
Can OCR read handwriting?
Neatly written block digits can be read to some extent, but cursive or messy writing won't reach a usable accuracy.
Carefully written block digits can be read to some extent, but cursive or messy writing won't reach a usable accuracy. Handwriting varies from person to person and strokes connect inconsistently, making it much harder to recognize than printed text. If you need to process lots of handwritten forms, consider a paid AI-based OCR service. This tool is best at quickly turning printed text shown on screen into text.
We do not guarantee the accuracy or completeness of the explanations on this page or the tool's output. OCR results may contain misreads, so always check important numbers and names yourself. Processing may be interrupted depending on your browser or connection. For details, see our Terms of Service and Privacy Policy.