Chinese Is Supported — But Not Bundled
Capture2Text can OCR Chinese well, but unlike Japanese, Chinese is not included by default. The seven bundled languages are English, French, German, Japanese, Korean, Russian and Spanish. For Chinese you install a language pack first — a two-minute job covered in our language installation guide. This page covers everything Chinese-specific.
Step 1 — Install the Chinese Language Pack
- Download
chi_sim.traineddata(Simplified) and/orchi_tra.traineddata(Traditional) from the SourceForge dictionaries section — links are in our download page’s language packs section — or from the tessdata GitHub repository. - Copy the
.traineddatafile(s) directly into the tessdata folder next toCapture2Text.exe. - Restart Capture2Text.
- Right-click the tray icon → OCR Language → select Chinese – Simplified or Chinese – Traditional.
chi_sim vs chi_tra: Which Do You Need?
| chi_sim (Simplified) | chi_tra (Traditional) | |
|---|---|---|
| Used in | Mainland China, Singapore | Taiwan, Hong Kong, Macau |
| Example | 简体中文 | 繁體中文 |
| Tip | Install both if you read mixed content; assign them to Win+1/Win+2 quick-access slots (Settings → OCR) to flip instantly. | |
Step 2 — Set Text Direction (Critical for Vertical Text)
Chinese appears both horizontally and vertically (book spines, traditional layouts, game UI). With a Chinese language active, press Win+O to cycle horizontal / vertical / auto. On auto, Capture2Text uses horizontal when the capture box is wider than twice its height, otherwise vertical. Wrong direction = garbage output, so check this before re-capturing.
Step 3 — Capture
Use the standard Win+Q box capture for paragraphs, or Win+E for single lines (see the hotkey reference). For best accuracy: capture at a readable size (zoom the source if the text is tiny), keep the box tight around the text, and enable Trim Capture in Settings → OCR.
Translating Chinese Captures
Open Settings → Translate tab to append a Google Translate rendering to your clipboard or popup alongside the OCR text. Remember this is the one feature that needs internet — the OCR itself stays offline. See managing languages for switching strategies, the Japanese OCR guide for CJK sibling notes, and all guides.
Frequently Asked Questions
Does Capture2Text support Chinese OCR?
Yes, but Chinese is not bundled — you must install it. Download chi_sim.traineddata (Simplified) and/or chi_tra.traineddata (Traditional), drop them into the tessdata folder, restart Capture2Text, and select the language from the tray menu. Steps below.
What is the difference between chi_sim and chi_tra?
chi_sim recognizes Simplified Chinese ( mainland China, Singapore), chi_tra recognizes Traditional Chinese (Taiwan, Hong Kong, Macau). Install both if you read mixed content, and switch between them via the tray menu or Win+1/2/3 quick-access keys.
How do I OCR vertical Chinese text?
Select chi_sim or chi_tra as the active language, then press Win+O to set text direction to vertical (or auto — auto picks vertical when the capture box is taller than twice its width). OCRing vertical text as horizontal produces garbage.
Why is my Chinese OCR output garbage characters?
Almost always one of two causes: the wrong language is selected (e.g. English active while capturing Chinese), or the text direction is wrong for vertical text. Fix both via the tray menu / Win+O and re-capture.
Where do I download the Chinese language pack?
From the dictionaries section of the Capture2Text SourceForge page (linked in our download page language packs section) or the tesseract-ocr/tessdata GitHub repository. You need chi_sim.traineddata and/or chi_tra.traineddata.
Can Capture2Text translate Chinese to English?
Yes — enable translation in Settings → Translate tab and the captured Chinese can be translated (via Google Translate) and appended to the clipboard or shown in the popup. Note: translation requires internet access; the OCR itself is offline.