English
This page is available in English
The whole page — instructions, buttons and results.
Switch to English→
No thanks, stay here
繁體中文
這個頁面有繁體中文版
整頁都是中文 —— 說明、按鈕和結果,都是。
前往繁體中文版→
不用了,留在這一頁
关于这个工具
查异体、简体、繁体与形近变体
查一个字的异体字、简繁与日本新旧字形,并看大陆、台湾、香港、日本和韩国的字表各收哪一种。数据来自 Unicode Unihan 18.0。
RT-TXT-056 · Text Tools
你的浏览器版本过旧,此工具的交互功能无法运行。请更新至 Safari 13.1、Chrome 或 Edge 80、Firefox 74 或以上版本。
异体字查询
输入汉字
Look up
Clear
Advertisement
After results · AD-W1
Responsive
同一个字,各地写法不同
「强」和「強」是同一个字,但大陆的字表收「强」,台湾、香港、日本和韩国的字表收「強」。「众」的繁体在台湾和香港是「眾」,在日本和韩国是「衆」。「广」在台湾是「廣」,在日本是「広」。这个工具把这些对照摆在一起:输入一个字,先看每个地区的字表收哪一种,再看其他写法。
怎么读一张卡片
同一个字的其他写法意义相同,一般可以互换。只在部分义项上相同就要小心:「着」和「著」只有部分用法相通。形近字只是长得像,是不同的字,会被用来仿冒网址。顺着写法再往下是间接连到的字:眾 → 𠂝 → 匝,这里每一步都只有一部字典的记录,越往下越要自己判断。
虚线框的字只有一部字典记录,也不算进各地字表的对照。「煙」的异体里有「完」,就只有 Lau 一部字典这么记,不要直接采用。
数据来自哪里
字形关系来自 Unicode 18.0 的 Unihan 数据库,共 16,825 个字。各地字表也由 Unihan 记录:大陆是《通用规范汉字表》(2013,8,105 字),香港是《常用字字形表》(2000),日本是《常用漢字表》(2010,2,136 字),韩国是教育用基础汉字(2007,1,800 字)。Unihan 没有记录台湾教育部的字表,所以台湾一栏用 Big5 的常用字区(5,401 字),这是一个近似。
📅 研究资料截至 2026年11月3日
· 来源: Unicode Han Database (Unihan) 18.0 and UAX #38, read 3 Nov 2026.
标准、已发表的公式与参考数据会随时间修订 — 依据这些数字行事前,请对照当前的权威来源核实。
Advertisement
After how-to · AD-W2
Responsive
常见问题
我该写哪一个?+
看读者在哪里。写给大陆读者用大陆字表里的字,写给台湾读者用台湾字表里的字。公文、证件和出版物要以当地现行规范为准,这里的对照只是起点。
不在表内就是不能用吗?+
不是。这些表只收常用字,人名和古籍里还有别的字形。例如「廣」不在日本的《常用漢字表》里,但在人名用汉字表里。
为什么有些字查不到异体?+
Unihan 不是异体字词典,只记录它整理过的关系。「青」和「靑」都有编码,但 Unihan 没有把它们连起来。查异体字,台湾教育部的《异体字字典》更全。
「同一个字的其他写法」真的能互换吗?+
Unicode 的定义是「一般可以替换」,不是处处可以。Unihan 记录字形关系,不判断哪个字形对。
输入的文字会上传吗?+
不会。第一次查询时浏览器下载一份字表,之后的查询都在浏览器里完成。
Related Tools
Rare Chinese Character Reader
Pinyin, Zhuyin, Cantonese and stroke count for rare characters
Text Tools
Try →
Chinese Character Encoding Lookup (Big5/GB/Unicode)
Unicode, UTF-8, UTF-16, Big5, GBK and GB18030 for any character
Developer Tools
Try →
Cangjie Input Code Lookup (倉頡)
Cangjie5 code for any character, with radical glyphs
Text Tools
Try →
Hanzi → Pinyin Converter (汉字拼音转换)
Hanzi → 5 pinyin formats + polyphone
Chinese Language & Script
Try →
方法与来源
计算方式
A lookup, not a calculation: 16,825 characters from Unicode Han Database 18.0, with their variant relations, the dictionaries that record each one, and the regional lists kTGH, kHKGlyph, kJoyoKanji, kKoreanEducationHanja and kBigFive level 1. The table is downloaded on the first lookup and read in the browser.
本工具依据的标准
Unicode Han Database variant relations as defined by UAX #38: kTraditionalVariant, kSimplifiedVariant, kSemanticVariant, kSpecializedSemanticVariant, kSpoofingVariant and kZVariant.
The relation TYPE is reported, not flattened into "variant". A spoofing variant and a traditional variant are different claims about two characters and conflating them would be misleading.
Unihan variant data is descriptive rather than prescriptive: it records relationships that exist in the writing system, not a ruling on which form is correct in a given jurisdiction.
Rebuilt 2026-11-03 in all three languages: the table was rebuilt from Unihan 18.0, keeping the source dictionaries and the Japanese new and old forms the earlier table dropped, and every result now comes from a tested kernel.
Which form each region lists comes from kTGH, kHKGlyph, kJoyoKanji and kKoreanEducationHanja. kUnihanCore2020 was measured and rejected as too wide: it puts 廣, 广 and 広 all in Japan's set. Unihan records no Taiwan education ministry list, so the Taiwan row uses Big5's frequently used level, and the page says that is an approximation.
A shape only one dictionary records is shown as such and kept out of the regional table, because such shapes are mostly loose pairs like 只 and 子.
资料来源
Unicode Han Database (Unihan) 18.0, © Unicode, Inc., under the Unicode Licence v3: https://www.unicode.org/Public/UCD/latest/ucd/Unihan.zip — read 2026-11-03.
UAX #38, Unicode Han Database (Unihan), for Unicode 18: https://www.unicode.org/reports/tr38/ — the definition of each variant relation and regional list, read 2026-11-03.
可能导致本页过时的因素
Unihan ships with each Unicode release and variant relations do change between versions. The vendored copy is a snapshot from 2026-06-17 and nothing refetches it, so a new Unicode version does not reach this page on its own.
Embed this tool
Put the 异体字查询 on your own site — free, no signup
Paste this where you want the tool to appear. It loads from
recatools.com, works on any site, sets no cookies, and shows no ads.
Please keep the credit line — it is the only thing we ask for.
Embed code for 异体字查询
Copy embed code
From our guides
Chinese character riddles, and how to solve them
一口咬掉牛尾巴 — one mouth bites off the ox's tail. Bite the tail off 牛, put 口 underneath, and you have writ...
Guide · 8 min read
Read →
How China, Taiwan and Unicode each choose between 峰 and 峯
峰 and 峯 are one word with one sound. Whether they count as one character depends on who is asking. Here is h...
Guide · 6 min read
Read →
Pick up where you left off
Clear
Stored only in this browser — never sent to our servers.