HanziTrace
Make your own Make one

About

We started HanziTrace because of a classroom moment that kept repeating: a student struggles with a character for weeks, then someone mentions that 家 is a roof over a pig — and the struggle is over in thirty seconds. The character was never hard. It just had never been introduced properly.

So this site does one thing: it introduces characters properly. Every character page gives you the stroke order, the meaning, and a story short enough to read in half a minute — then a free practice sheet, 10 ready-made packs and a generator for whatever you need this week, so your hand can learn what your head just did. Everything is free, there is nothing to sign up for, and you may print as many copies as you like.

We are a small operation and we say so. The stories are drafted with AI help and then read by a person, one by one; the ones nobody has read yet are marked as drafts on their own pages, and the counts below are live. If you find a mistake, an email genuinely changes the site — the corrections readers send are how the notes get better.

Who it is for

Mandarin teachers, weekend Chinese schools, and bilingual families — people who need a page a child can write on tonight. Adult beginners use it too, but the packs are built and ordered the way a class or a parent would want them.

What has been checked, and what has not

This is the part worth reading before you put a sheet in front of a class.

  • Stroke counts and stroke order come from Make Me a Hanzi, and every stroke count in our database was compared character by character against that source with no disagreements.
  • English glosses come from CC-CEDICT. A dictionary's first entry is not always the sense a beginner needs, so worksheets print a shortened gloss where we have one. 212 characters are flagged in our own data as needing a person to write a better one; until that happens they print the dictionary's wording.
  • Pinyin is per-character in our database, which is wrong the moment two characters make a word — 行 is xíng alone and háng in 银行. The generator reads your text with a word-aware engine, lists every character that has more than one reading, and lets you change any of them. It does not decide silently.
  • Culture notes: 253 characters have one, and 53 of those have been read by a person. The remaining 200 are AI drafts and are marked as drafts on their own pages.
  • Teacher's notes on the packs are written from general teaching practice. None of them has been reviewed by a classroom teacher yet.

Why a note says what kind it is

A story about a character can be three different things, and they are not equally true.

Where it comes from 字源 0 characters
The character’s own history, with a source.
A way to remember it 记忆联想 241 characters
A picture that makes the shape stick — not a claim about the character’s history.
How it is used 文化习俗 12 characters
Custom and usage, in the settings where it applies.

The strongest of the three is the one we are strictest about: a note is only presented as the character's history when a citable source is attached to it, which is why that count is currently 0. 茶 is a good example of the difference. "Grass on top, a tree underneath, a person picking in between" is an excellent way to remember the shape, and it is not where the character came from — 茶 is a later form of 荼. So it is labelled as a memory aid, on the web page and on the printed sheet.

Where the data comes from

See using these sheets for what the licences mean for you.

Telling us something is wrong

Write to [email protected]. A wrong reading, a note that overstates itself, or a sheet that printed badly on your printer are all worth an email — those are the reports that change the site.

HanziTrace is made by hanzitrace.