Open data

Korean word frequency list

80,000 words · CC BY 4.0 · 2026

The 80,000 most frequent Korean words in their dictionary forms, counted by Shiberu from web pages, online comments, film and TV subtitles and Wikipedia. It is the list behind the frequency ranks in the Shiberu Reader and browser extension, and it is free for anyone to use, commercially too.

TSV · 1.8 MB

TSV, gzip · 0.7 MB

README

Licence text

UTF-8, tab-separated, with a header row and three columns: rank (1 is the most frequent), lemma (the dictionary form) and per_million (occurrences per million words).

Stable links that always point at the newest version: ko-frequency-latest.tsv and ko-frequency-latest.tsv.gz.

Look up a word

Word

Type the dictionary form of the word.

The 50 most frequent words

Rank

Word

Per million

1

하다

28,341

2

있다

22,879

3

되다

10,749

4

보다

9,650

5

없다

8,714

6

아니다

6,619

7

가다

6,575

8

이

6,546

9

수

6,398

10

것

6,201

11

그

5,384

12

사람

4,833

13

않다

4,652

14

같다

4,612

15

알다

4,278

16

등

3,973

17

왜

3,816

18

다

3,544

19

좋다

3,393

20

주다

3,215

21

그렇다

3,128

22

말

3,042

23

모르다

2,912

24

뭐

2,864

25

이다

2,846

26

더

2,837

27

말다

2,807

28

한

2,674

29

받다

2,667

30

대하다

2,647

31

위하다

2,525

32

지금

2,444

33

저

2,402

34

및

2,378

35

좀

2,177

36

맞다

2,172

37

많다

2,128

38

오다

2,110

39

우리

2,106

40

때

2,104

41

중

2,030

42

그냥

1,947

43

죽다

1,873

44

만들다

1,854

45

내

1,803

46

전

1,787

47

나다

1,778

48

보이다

1,763

49

정말

1,747

50

무슨

1,731

How it was made

Counted from these sources, with these shares of the blend:

FineWeb-2 (kor_Hang)

17.5%

Web pages · ODC-By 1.0; Common Crawl terms of use

CC-100 Korean

17.5%

Web pages · No claims on the preparation; Common Crawl terms of use

Naver news comments (KcBERT corpus), DC Inside comments, NSMC movie reviews

25%

Online comments · CC BY-SA 4.0; Apache-2.0; CC0 1.0

OpenSubtitles v2024 (OPUS)

25%

Film and TV subtitles · OPUS: credit opensubtitles.org and cite Lison & Tiedemann 2016

Korean Wikipedia

15%

Encyclopedia · CC BY-SA 4.0

Every text went through Shiberu's Korean analyzer (Kiwi, with the same word grouping and dictionary-form rule the Reader and the dictionary popup use), so each entry is a word as a learner looks it up: 있는 → 있다, 교무실에서 → 교무실, 공부했어요 → 공부하다.

Each source gives a rate per million words for every word. The blend is a weighted median: a word's score is the highest rate that sources holding at least half of the weight reach. The written side (web and Wikipedia) or the informal side (comments and subtitles) can carry a word, but no single source can lift its own oddities, such as Wikipedia's citation words or character names in subtitles. A word must appear in at least two sources.

Left out: numbers and anything with a digit, words written only in Latin letters or hanja, anything with a bare jamo (ㅋㅋ, ㅎㅎ, ㅠㅠ), and particles that punctuation cut off their noun.

Only counts are published: the files contain no text from any source.

How it compares

Against the CC100-Korean list (counted in mecab-ko morphemes), the Spearman rank correlation over the words both lists rank in their top 20,000 is 0.746 (11,006 shared words); the older subtitles-only list Shiberu used before scores 0.513. Most disagreements come from how words are cut: CC100-Korean ranks particles, endings and split nouns (it counts 선생 and 님 apart, so 선생님 is no. 59,854 there). Spoken words rank higher here, because subtitles and comments are half of the weight.

Word

This list

CC100-Korean

오가다

3,727

3,894

몰려들다

11,205

10,666

따돌리다

12,754

8,976

왜

17

518

뭐

24

605

그냥

42

796

제발

164

4,036

선생님

926

59,854

Licence and credit

Free to copy, share and adapt for any purpose, commercial use included, under the Creative Commons Attribution 4.0 licence. The one condition is credit: please name Shiberu with this line.

Shiberu Korean Word Frequency List 2026 by Shiberu (https://shiberu.app/korean-frequency-list), licensed under CC BY 4.0.

When you share the files themselves, keep the README with them: it credits the sources as their licences ask. The list holds statistics about the texts (which words occur and how often), not the texts.

Learning Japanese too? See the Japanese frequency list.

Questions or corrections: [email protected].