Arena 직군별 종합

이 벤치마크는?

전문가 수준의 프롬프트에서의 사용자 선호도를 집계한 Arena Elo 점수입니다.

출처: Arena Intelligence
순위모델
#1

Anthropic

Claude Opus 4.6

1538
#2

OpenAI

GPT-5.4

1526
#3

OpenAI

GPT-5.4 Pro

1526
#4

Google

Gemini 3.1 Pro

1516
#5

Xiaomi

MiMo-V2-Pro

1506
#6

Anthropic

Claude Opus 4.5

1504
#7

Anthropic

Claude Sonnet 4.6

1503
#8

Anthropic

Claude Sonnet 4.5

1498
#9

Google

Gemini 3 Flash

1496
#10

OpenAI

GPT-5.4 Mini

1495
#11

Moonshot AI

Kimi K2.5

1493
#12

Google

Gemma 4 31B

1488
#13

Z.ai

GLM-5.1

1485
#14

Z.ai

GLM-5

1484
#15

Anthropic

Claude Opus 4.1

1483
#16

Grok

Grok 4.20

1482
#17

Grok

Grok 4.20 (Reasoning)

1482
#18

Google

Gemini 2.5 Pro

1466
#19

Qwen

Qwen3.5 397B A17B

1463
#20

OpenAI

GPT-5.4 Nano

1462
#21

OpenAI

GPT-5

1460
#22

Meituan

Longcat Flash Chat

1458
#23

Baidu

ERNIE 5.0 Thinking

1453
#24

Google

Gemini 3.1 Flash Lite

1450
#25

MiniMax

MiniMax M2.5

1448
#26

Anthropic

Claude Opus 4

1448
#27

DeepSeek

DeepSeek V3.2

1448
#28

Anthropic

Claude Haiku 4.5

1445
#29

Grok

Grok 4.1 Fast

1438
#30

Grok

Grok 4.1 Fast (Reasoning)

1438
#31

Anthropic

Claude Sonnet 4

1435
#32

Google

Gemini 2.5 Flash

1428
#33

MiniMax

MiniMax M2.7

1422
#34

Arcee AI

Trinity Large Thinking

1414
#35

OpenAI

GPT-5 Mini

1404
#36

Google

Gemini 2.5 Flash Lite

1387
#37

Amazon

Nova 2 Lite

1365
#38

OpenAI

GPT OSS 120B

1362
#39

OpenAI

GPT-5 Nano

1355
#40

Meta AI

Llama 4 Maverick

1324
#41

Meta AI

Llama 4 Scout

1307
#42

OpenAI

GPT-4.1

1286