JIS Levels and Kuten Codes: Japan's Character Classification System
How Japan classifies kanji into 4 levels across JIS X 0208 and JIS X 0213, with kuten positional codes.
The 94×94 Grid
JIS character sets are organized as a grid of 94 rows (区 ku) and 94 columns (点 ten). Each character is identified by its row-column position, called a kuten code.
For example, 亜 is at row 16, column 1 — written as 1面16区1点 (plane 1, row 16, cell 1). JIS X 0213 added a second plane, so characters can be at 1面-XX-XX or 2面-XX-XX.
The Four JIS Levels
Kanji in JIS standards are classified by frequency and importance into four levels:
| Level | Standard | Count | Content |
|---|---|---|---|
| Level 1 (第一水準) | JIS X 0208 | 2,965 | Common everyday kanji |
| Level 2 (第二水準) | JIS X 0208 | 3,390 | Less common but essential |
| Level 3 (第三水準) | JIS X 0213 | 1,259 | Names, historical text |
| Level 4 (第四水準) | JIS X 0213 | 2,436 | Rare kanji, supplements |
Levels 1-2 are from JIS X 0208 (1978/1983/1990/1997). Levels 3-4 were added by JIS X 0213 (2000/2004) to cover characters needed for personal names and historical documents.
JIS X 0208 vs JIS X 0213
JIS X 0213 is not just an extension — it also revised some character assignments from JIS X 0208. The 2004 revision added 10 new characters and changed the example glyphs for 168 existing characters.
| Character | JIS X 0208 | JIS X 0213:2004 |
|---|---|---|
| 繋 (U+7E4B) | Level 1, 1-23-50 | Same position |
| 繫 (U+7E6B) | Not included | Level 3, added at 1-94-94 |
Both forms of the character were recognized as valid, but they map to different Unicode code points. This tool shows the JIS level and kuten code for each.
Practical Significance
JIS levels matter for Japanese IT systems:
- Government systems often require Level 1-2 support at minimum
- Name processing (koseki/jūminhyō) needs Level 3-4 for rare name kanji
- Font requirements: Level 1-2 fonts are common, Level 3-4 fonts are specialized
- Input methods: Some IMEs only support Level 1-2 by default
Related articles
Han Unification: How Unicode Merged 100,000 CJK Characters
How the IRG decided which characters from Japan, China, Taiwan, and Korea are 'the same,' with a tool to check any character's source.
IVS: How Unicode Represents 47 Versions of the Same Kanji
Understanding Ideographic Variation Sequences and Standardized Variation Sequences, with live font rendering of all registered variants.
Why One Font Isn't Enough: CJK Variant Coverage Across Fonts
How different CJK fonts implement different IVD collections, why a single font can't show every registered variant, and how this site combines three fonts to render every IVS faithfully.