How many kanji are there
The general-use list has 2,136 kanji. Elementary school adds them 80, 160, 200, 202, 193, 191 — 1,026 by the end of grade 6.
The number to learn toward, if you are learning the Japanese that newspapers and textbooks agree on, is 2,136. That is the jōyō kanji list, the general-use list in the Cabinet notification of 30 November 2010. It is not every character that has ever been written in Japan. Names use a further name-kanji list. Specialist writing uses more. “How many kanji are there” as a question about the universe of characters is the same trap as the Unicode total on the Chinese side. As a question about what school expects, it has a finite answer.
1,026 of them are the elementary list
The elementary subset is 1,026 kanji, taught across six grades. The allocation is 80, then 160, then 200, then 202, then 193, then 191. Those six numbers sum to 1,026. The other 1,110 characters in the jōyō list are the secondary-school remainder, not a rival list with a different name.
[[chart:ld-kanji-grades]]The first view is what gets added each year, so grade 1 does not disappear next to the secondary block. Grade 1 is 80 characters. That is a term of work, not a lifetime. Switch to the running total: 80, 240, 440, 642, 835, 1,026 at the end of elementary, 2,136 for the full general-use list. The last jump is the 1,110. It looks large because it is six years of elementary work arriving as one bar. It is still a list, not an open set.
What this chart is not
It is not a JLPT chart. The JLPT does not publish an official kanji list, and unofficial “N5 = 100 kanji” tables disagree with each other. Use them as someone else’s guess. Use the grade table if you want a sequence a country actually teaches. When you are ready to try a timed practice test, choose JLPT — a timed paper, not a substitute for the 80.
It is also not the Chinese common set. China’s level 1 is 3,500 characters and is not “grade 1”. Putting 80 and 3,500 in one sentence without that caveat makes both countries sound arbitrary. The comparison, with the jobs labelled, is the three lists.
How to start without a scar
“How did kanji get his scar” is a different Kanji. The character 漢字 does not have a plot. Start with the 80. For each one: the character, one on-reading, one kun-reading if it has one, and one word. Cover the readings. Write the character from the word, then say the word from the character. A new pile of 160 in grade 2 is what you add after the 80 are quiet, not on the same afternoon.
Why this method works
Eighty characters have an end. “All kanji” does not. A closed grade list lets you keep testing the ones you already recalled once, which is the schedule that held up a week later in the vocabulary study — not because that study used kanji, but because an open list tempts you to drop anything that felt familiar and hunt a new character instead.
Run this session
Twenty-five minutes, grade 1 only. If you are past those 80, use the next grade’s first 15. Do not touch the secondary 1,110.
- 0–3 min. From memory, how many kanji in grade 1, and how many in the full jōyō list? 80 and 2,136. Check the cumulative view if you were off.
- 3–20 min. Fifteen grade-1 kanji you cannot write cold. One card each: character, one on-reading, one kun-reading if the list gives one, one word. Cover the readings. Write the character from the word. Hits stay in the pile for another pass. Misses get a second try at the end, not a new character.
- 20–25 min. Count the pile. If you added a 16th because the first 15 felt smooth, take it back out. Smooth is the five-minute trap.
Common traps
- A downloaded “JLPT N5 kanji” list with a confident count. Fix: the JLPT does not publish that list. Use the grade table.
- Mixing Japan’s 80 with China’s 3,500 as if both were “first year”. Fix: they are different countries’ lists. See the three-list chart.
- Studying name kanji before the 80. Fix: names are a separate list. They can wait until a name is in front of you.
Rate the pile in Study Lab. The reason rereading the list feels faster, and fails a week later, is the retrieval result, measured on vocabulary pairs rather than on a kanji product trial.