Chinese character extractor
Keep only Chinese characters and drop punctuation, digits, Latin letters and kana: one character per line by default for counting, or one run per line with the merge option; rare extension-block characters can be included too.
Runs in your browserEvery computation happens in your browser — your data never leaves this device.
Extracted characters
One item per line, ready to copy.
Paste text to extract Chinese characters
Matches: 0 · Unique: 0
What this tool does
- Clean mixed Chinese/English/Japanese text down to pure Han characters for typography or glyph review.
- Count Chinese characters: one per line by default, and the stats line gives the total.
- Strip digits and Latin noise out of OCR output before comparing keywords.
Example
Input
Hello 世界，UniKit 工具站 2026：中文 123 と日本語です。
Output
世界 工具站 中文 日本語
Japanese kanji such as 日本語 count as Han characters, while kana, punctuation and digits are dropped; with merge enabled each run of characters is one line.
Frequently asked questions
Do Japanese kanji count as Han characters?
Yes — classification follows the Unicode Han script, so 日本語 is extracted while kana such as と or です is not.
Why one character per line by default?
It makes proofreading and counting trivial. Turn on “Merge consecutive characters” to get one run per line instead.
Are punctuation and digits kept?
No. Chinese and Latin punctuation, digits, letters and emoji are all dropped; only Han characters survive.
Are rare characters supported?
The default covers the basic block, extension A and the compatibility block. Enable “Include rare extension characters” to also pick up extension B and beyond.
How much text can it handle?
Two million characters, after which it reports an error rather than freezing the main thread.
Keywords：汉字提取中文提取hanzi extractor只保留中文中文汉字提取汉字中文清洗汉字计数