fedi-index: Language: 한국어

Best models for this language by Arena votes in it. Provisional: one source family.

Live data for 2026-10 (recalculated daily) · Methodology v1 · Methodology · History · JSON

All Open weights Closed

#

Tier

Model

fedi-index

Percentile

Value

Sources

11

A

1,459 ±17

93

+44

A

14

A

1,451 ±15

90

+45

A

18

A

1,443 ±13

87

+80

A

23

A

1,441 ±21

84

+105◆

A

24

A

1,437 ±16

83

+44

A

28

A

1,432 ±13

80

+36

A

29

B

1,432 ±17

79

+23

A

35

B

1,427 ±12

75

+36

A

43

B

1,419 ±16

69

+27

A

46

B

1,416 ±17

67

+26

A

51

B

1,412 ±23

63

+46

A

54

B

1,409 ±23

61

+27

A

57

B

1,406 ±12

58

+27

A

58

B

1,406 ±15

58

+71

A

62

B

1,404 ±17

55

+18

A

65

B

1,401 ±14

53

+50

A

71

C

1,393 ±15

48

+75

A

72

C

1,391 ±19

47

+32

A

75

C

1,388 ±15

45

+55◆

A

77

C

1,386 ±20

44

+14

A

78

C

1,385 ±19

43

+38

A

80

C

1,384 ±19

42

+11

A

82

C

1,384 ±13

40

+50

A

83

C

1,382 ±19

39

+14

A

88

C

1,370 ±22

36

+19

A

89

C

1,369 ±17

35

+25

A

92

C

1,360 ±24

33

-22

A

94

C

1,358 ±12

31

+32

A

96

C

1,352 ±16

30

-2

A

98

C

1,351 ±18

28

-22

A

99

C

1,350 ±22

27

-10

A

100

C

1,350 ±18

27

-30

A

105

D

1,335 ±17

23

-2

A

106

D

1,321 ±17

22

-36

A

107

D

1,319 ±19

22

-29

A

108

D

1,317 ±22

21

-14

A

110

D

1,317 ±18

19

—

A

111

D

1,300 ±17

18

+9

A

124

D

1,176 ±15

9

-236

A

AArena (human votes) ◆ Pareto frontier: nothing is both better and cheaper

fedi-index is an Elo-equivalent on a fixed scale: comparable between months within one methodology version. Tiers S–D: by percentile in the slice, and no better than the confidence interval allows (cut-offs in the methodology). This slice has one source family, so all its scores are provisional. Methodology