OpenCC (Open Chinese Convert) is an open-source project for converting between Chinese character standards. It is a dictionary engine, not a translator and not a spellchecker. ChineseDocConverter uses those conversion types in the text and document tools so you can pick a source and a target instead of hoping a single "Chinese" button is enough.
The names in the dropdown match common OpenCC configurations. If you have used the command-line OpenCC tool, the idea is the same: a chain of dictionaries that first identify the source standard, then write the target standard, and optionally apply phrase replacements.
What each option is for
| Option in the app | Typical use | What it does not do |
|---|---|---|
| Simplified → Traditional | Generic Traditional output when the region is still unknown | It does not pick Taiwan or Hong Kong spellings for you |
| Simplified → Traditional (Taiwan) | Documents aimed at readers in Taiwan | It does not rewrite the article into Taiwan Mandarin style |
| Simplified → Traditional (Taiwan, with phrases) | Taiwan characters plus some wording | It is still not a localization service |
| Simplified → Traditional (Hong Kong) | Documents aimed at Hong Kong or similar written practice | It does not turn the text into spoken Cantonese |
| Traditional → Simplified | Generic Simplified output | It does not recover a lost encoding |
| Traditional → Traditional (Taiwan) | Normalize already-Traditional text toward Taiwan | It does not simplify strokes |
| Traditional → Traditional (Hong Kong) | Normalize already-Traditional text toward Hong Kong | It does not simplify strokes |
| Traditional (Taiwan) → Simplified | A known Taiwan source going to Simplified readers | It does not translate into English |
| Traditional (Taiwan) → Simplified (with phrases) | Taiwan source plus phrase tables | It does not make the prose sound like a Beijing newspaper |
| Traditional (Hong Kong) → Simplified | A known Hong Kong source going to Simplified readers | It does not remove Cantonese grammar if the author wrote it |
If two options look similar, prefer the one that names the region you actually have. "Generic" exists for mixed or unknown files. It is a starting point, not a gold standard.
Characters vs phrases
A character conversion changes 门 to 門, 国 to 國, 书 to 書. That layer is what most people mean by Simplified versus Traditional.
A phrase table may also change a regional word, such as a software term or a daily-life noun that Taiwan and mainland China do not share. Phrase types are useful for user interfaces and ordinary news prose. They are still not a translator. They will not decide whether a joke works, whether a legal disclaimer is valid in another jurisdiction, or whether a product name should stay in its original characters.
If you need only glyphs, pick a type without phrases. If you need a slightly more local word list, pick a phrase type, then read the output. Phrase tables can be surprising in names that happen to contain a common word.
How OpenCC decides
OpenCC does not "understand" a sentence the way a person does. It matches the longest dictionary entries it can, then falls back to single characters. That is why 头发 has a good chance of becoming 頭髮: the two-character phrase is in the table. A rare personal name that contains 发 has a weaker chance of being right.
This design is also why converting a file twice can drift. The second pass sees the first pass's choices, not the author's original intent. Keep the source file. Run one conversion for one audience.
Context windows help on common mergers such as 后 / 後, but they fail on short labels. A table cell that only says 后 gives the engine almost nothing to work with. One-to-many mappings lists the usual problem characters.
What OpenCC is not
- It is not handwriting recognition or OCR. Pictures of characters stay pictures.
- It is not Japanese conversion. Shinjitai and kyūjitai are a different history, even when a kanji looks familiar.
- It is not Korean hanja conversion.
- It is not a grammar checker and not a pinyin annotator. For pronunciation spelling, use the Pinyin converter.
- It is not a political statement about which standard is "real Chinese." It is a mapping between published standards.
For the educational background of the two systems, read Simplified vs Traditional Chinese. For the methods used in the twentieth-century reforms, read how characters were simplified.
Choosing an option in practice
- Name the reader: mainland China, Singapore, Taiwan, Hong Kong, or mixed.
- Name the source if you know it. A Taiwan file going to Simplified is not the same job as a Hong Kong file going to Simplified.
- Decide whether you want phrase substitution.
- Convert a sample, then the whole document.
- Proofread names, titles and legal lines.
If you are comparing Taiwan and Hong Kong output, paste the same paragraph into the text converter twice. The differences are easier to see than to memorize.
Frequently asked questions
Which type should a global company pick as default? There is no single default that satisfies Taiwan and Hong Kong. Many products keep zh-CN, zh-TW and zh-HK as separate locales.
Why is there Traditional to Traditional? Because a file can already be Traditional and still use the other region's preferred glyphs. Normalization is a real publishing task.
Do I need OpenCC if my word processor has a Chinese convert button? Office buttons vary in quality and rarely expose Taiwan versus Hong Kong the same way. They also tend to convert a selection, not the hidden parts of a document. This site converts the file so styles stay in place.
Can I add my own words? The public web tools use the standard OpenCC dictionaries. They do not load a private company glossary. If you have a protected term, search for it after conversion.
Will OpenCC fix mojibake? No. Repair garbled Chinese text first.