Arena 창작 Elo

이 벤치마크는?

창작 글쓰기 프롬프트에서의 사용자 선호도를 집계한 Arena Elo 점수입니다.

출처: Arena Intelligence
순위모델
#1

Anthropic

Claude Opus 4.6

1491
#2

Google

Gemini 3.1 Pro

1488
#3

Grok

Grok 4.20

1480
#4

Grok

Grok 4.20 (Reasoning)

1480
#5

Anthropic

Claude Opus 4.5

1469
#6

Google

Gemini 3 Flash

1461
#7

OpenAI

GPT-5.4

1459
#8

OpenAI

GPT-5.4 Pro

1459
#9

Z.ai

GLM-5.1

1452
#10

Google

Gemini 2.5 Pro

1449
#11

Z.ai

GLM-5

1445
#12

Anthropic

Claude Opus 4.1

1445
#13

Anthropic

Claude Sonnet 4.6

1444
#14

Anthropic

Claude Sonnet 4.5

1441
#15

Anthropic

Claude Opus 4

1431
#16

Google

Gemma 4 31B

1429
#17

Xiaomi

MiMo-V2-Pro

1426
#18

Baidu

ERNIE 5.0 Thinking

1424
#19

Moonshot AI

Kimi K2.5

1420
#20

Qwen

Qwen3.5 397B A17B

1418
#21

OpenAI

GPT-5.4 Mini

1417
#22

Google

Gemini 3.1 Flash Lite

1414
#23

Grok

Grok 4.1 Fast

1412
#24

Grok

Grok 4.1 Fast (Reasoning)

1412
#25

DeepSeek

DeepSeek V3.2

1402
#26

Google

Gemini 2.5 Flash

1398
#27

Anthropic

Claude Sonnet 4

1395
#28

Meituan

Longcat Flash Chat

1390
#29

Anthropic

Claude Haiku 4.5

1384
#30

MiniMax

MiniMax M2.5

1380
#31

OpenAI

GPT-5

1375
#32

Google

Gemini 2.5 Flash Lite

1361
#33

Arcee AI

Trinity Large Thinking

1357
#34

MiniMax

MiniMax M2.7

1345
#35

OpenAI

GPT-5.4 Nano

1340
#36

OpenAI

GPT-5 Mini

1326
#37

Meta AI

Llama 4 Maverick

1308
#38

Meta AI

Llama 4 Scout

1290
#39

OpenAI

GPT-4.1

1286
#40

OpenAI

GPT OSS 120B

1280
#41

Amazon

Nova 2 Lite

1277
#42

OpenAI

GPT-5 Nano

1250