AA Intelligence Index

이 벤치마크는?

Artificial Analysis가 산출하는 종합 지능 지수. 추론, 코딩, 수학, 언어 이해 등 다양한 벤치마크를 가중 평균하여 AI 모델의 전반적인 역량을 하나의 수치로 표현합니다.

출처: Artificial Analysis
순위모델
#1

Google

Gemini 3.1 Pro

57.2
#2

OpenAI

GPT-5.4

56.8
#3

Anthropic

Claude Opus 4.6

53.0
#4

Meta AI

Muse Spark

52.1
#5

Qwen

Qwen3.6 Plus

50.0
#6

Z.ai

GLM-5

49.8
#7

Anthropic

Claude Opus 4.5

49.7
#8

MiniMax

MiniMax M2.7

49.6
#9

Grok

Grok 4.20 (Reasoning)

49.3
#10

Xiaomi

MiMo-V2-Pro

49.2
#11

Moonshot AI

Kimi K2.5

46.8
#12

Google

Gemini 3 Flash

46.4
#13

Anthropic

Claude Sonnet 4.5

43.0
#14

Anthropic

Claude Sonnet 4.6

42.6
#15

Anthropic

Claude Opus 4.1

42.0
#16

MiniMax

MiniMax M2.5

41.9
#17

DeepSeek

DeepSeek V3.2

41.7
#18

Qwen

Qwen3.5 397B A17B

40.1
#19

Google

Gemma 4 31B

39.2
#20

Anthropic

Claude Opus 4

39.0
#21

Anthropic

Claude Sonnet 4

38.7
#22

Grok

Grok 4.1 Fast (Reasoning)

38.6
#23

OpenAI

GPT-5.4 Nano

38.1
#24

OpenAI

GPT-5.4 Mini

37.7
#25

Anthropic

Claude Haiku 4.5

37.1
#26

NVIDIA

Nemotron 3 Super

36.0
#27

Google

Gemini 2.5 Pro

34.6
#28

Google

Gemini 3.1 Flash Lite

33.5
#29

Baidu

ERNIE 5.0 Thinking

29.1
#30

Grok

Grok 4.20

29.0
#31

Google

Gemini 2.5 Flash

27.0
#32

OpenAI

GPT-4.1

26.3
#33

OpenAI

GPT-5 Nano

25.9
#34

OpenAI

GPT OSS 120B

24.5
#35

Meituan

Longcat Flash Chat

23.9
#36

Grok

Grok 4.1 Fast

23.6
#37

OpenAI

GPT-5

21.8
#38

OpenAI

GPT-5 Mini

20.7
#39

Google

Gemini 2.5 Flash Lite

19.4
#40

Meta AI

Llama 4 Maverick

18.4
#41

Amazon

Nova 2 Lite

18.0
#42

Baidu

ERNIE 4.5 300B A47B

15.0
#43

Meta AI

Llama 4 Scout

13.5
#44

Mistral AI

Mistral Small 4

9.0