Capture2Text: The Fastest Free Offline OCR for Windows
Instantly extract text from any part of your screen with a simple keyboard shortcut. No internet required. 90+ languages supported including Chinese, Japanese, and Korean.
Why Choose Capture2Text?
More than just a standard OCR tool. It’s a productivity powerhouse designed for speed, privacy, and multilingual support.
Ultra-Fast Text Extraction
The fastest workflow in the market. Press Win + Q, select your area, and the text is already in your clipboard. No manual copying needed.
90+ Languages Supported
Optimized for complex scripts. Best-in-class support for Chinese (Simplified/Traditional), Japanese, and Korean characters.
100% Private & Offline
Unlike cloud-based OCR tools, Capture2Text processes everything locally on your PC. No data is ever sent to any server. Perfect for sensitive documents.
One-Click Translation
Automatically translate recognized text. Supports integration with Google Translate to show instant translation in a popup balloon.
Full Hotkey Control
Fully customizable keyboard shortcuts. Rebind capture, translation, and text-to-speech triggers to fit your personal workflow.
Voice Recognition
Not just for images! Capture2Text includes an experimental speech-to-text feature that recognizes and translates spoken audio.
How to Use Capture2Text in 3 Easy Steps
Master OCR extraction in seconds. No complex setup required.
Launch & Setup
Run the Capture2Text.exe. Right-click the tray icon to select your target OCR language.
Select Area
Press Win + Q. Click and drag your mouse to box the text you want to capture.
Instant Result
Release the mouse. The text is copied to your clipboard and shown in a popup window.
Common Real-World Scenarios
Manga & Comic Translation
Easily capture vertical text in Japanese Manga. Use the “Speech Bubble” mode for better accuracy in comic dialogue boxes.
Game Text Capture
Translate non-copyable text from video games in real-time. Perfect for visual novels and JRPGs.
Data Entry from PDFs
Quickly pull data from scanned PDF documents, protected images, or YouTube video frames into Excel or Word.
Coding & Technical Support
Copy error messages from dialogue boxes or system tooltips that don’t allow standard text selection.
Frequently Asked Questions
Everything you need to know about using Capture2Text on Windows.
Q: Is Capture2Text really free to use?
Yes, Capture2Text is 100% free and open-source software licensed under the GNU General Public License (GPL) v3. There are no hidden fees, subscriptions, or premium tiers.
Q: Does it support Windows 11 and Windows 10?
Absolutely. While the official version was released a few years ago, it remains fully compatible with Windows 11 and all versions of Windows 10. No special compatibility mode is required.
Q: How do I improve the OCR accuracy for Chinese or Japanese?
Accuracy can be improved by ensuring you have the latest Tesseract language data. You can also adjust the “OCR Scale” in the settings. For Japanese Manga, make sure to enable the vertical text recognition option in the tray icon settings.
Q: Why is my hotkey (Win + Q) not working?
This is usually caused by hotkey conflicts with other software (like Windows Search or QuickLook). You can easily change the capture shortcut to something else, like Ctrl + Q, by right-clicking the Capture2Text icon and selecting “Settings”.
Technical Specifications
Optimizing For Better Results
To get the highest OCR recognition rate, we recommend adjusting these three critical settings in the Capture2Text options panel:
- 1. OCR Scale Factor If the text is small, set the OCR Scale to 3.0 or 3.5. This effectively “zooms in” on the text before processing, dramatically improving accuracy.
- 2. Whitelist Characters Only need numbers? Use the whitelist feature to limit the engine’s focus, preventing “0” from being mistaken for “O”.
- 3. Deskewing Images Enable automatic deskewing in settings to handle text that is tilted or slightly rotated on your screen.
How Capture2Text Compares to Other OCR Tools
Why thousands of users still prefer this open-source tool over modern alternatives.
| Feature | Capture2Text | ShareX (OCR) | Windows Snipping Tool | Online OCR APIs |
|---|---|---|---|---|
| Offline Processing | ✅ Native (Local) | ❌ Cloud Required | ✅ Local | ❌ Internet Only |
| Hotkeys (Speed) | ✅ Instant (Win+Q) | Average | Slower (Multi-click) | ❌ No Hotkeys |
| Language Support | ✅ 90+ (Tesseract) | Limited to OS | System Only | Varies |
| Vertical Text (JP/CN) | ✅ Specialized Mode | ❌ Poor Support | ❌ Basic Only | Manual Only |
| Privacy / Security | ⭐⭐⭐⭐⭐ | ⭐⭐⭐ | ⭐⭐⭐⭐ | ⭐ |
Verdict: While tools like ShareX are great for screenshots, Capture2Text remains the superior choice for heavy-duty OCR tasks, specialized languages (Manga/Visual Novels), and users who demand absolute data privacy.
Common Troubleshooting
“Tesseract Data Not Found” Error
This usually happens when the tessdata folder is missing or incorrectly named. Ensure that your language files (e.g., jpn.traineddata) are placed inside the Capture2Text/tessdata/ directory. Do not subfolder them.
Application Fails to Launch (Side-by-Side Error)
This is a classic Windows dependency issue. You must install the Microsoft Visual C++ 2015 Redistributable. Without this, the OCR engine cannot initialize on Windows 10/11.
Inaccurate Results on High-DPI Screens
If you use 4K monitors or 150% scaling, right-click Capture2Text.exe > Properties > Compatibility > Change high DPI settings > Override high DPI scaling behavior and set it to “System (Enhanced)”.
🚀 Pro Tips for Speed
- 💡 Output to Clipboard: Enable “Save to Clipboard” in settings to bypass the popup window for a 100% silent workflow.
- 💡 Use Whitelists: If you’re only capturing numbers (like in Excel), set a whitelist to
0123456789to eliminate letter-number confusion. - 💡 Quick Switch: Use the
Win + Oshortcut to quickly cycle through your pre-set OCR languages without opening the menu. - 💡 Background Contrast: For dark-mode apps, try toggling “Negative OCR” in settings if the standard recognition fails.
Open Source & Free
Licensed under GNU GPL v3
A Community-Driven Project
Capture2Text is an open-source project originally created by Christopher Brochtrup. This website is a community resource dedicated to providing optimized builds, easy-to-read documentation, and technical support for users worldwide.
Our Commitment to Transparency:
- No Bloatware: We only provide clean, portable ZIP files identical to the source or pre-bundled with official Tesseract data.
- Respect the Source: We always link back to the official SourceForge repository for the latest source code.
- Privacy First: No tracking, no account required, and no data collection. Just pure OCR utility.