LCR (Long Context Recall)

이 벤치마크는?

긴 문서 내에서 특정 정보를 정확히 찾아내는 능력을 평가하는 벤치마크. 모델의 긴 컨텍스트 처리 및 검색 능력을 측정합니다. 점수는 정답률(%)입니다.

출처: Artificial Analysis
순위모델
#1

Anthropic

Claude Opus 4.5

74.0%
#2

OpenAI

GPT-5.4

74.0%
#3

Google

Gemini 3.1 Pro

72.7%
#4

Anthropic

Claude Opus 4.6

70.7%
#5

Anthropic

Claude Haiku 4.5

70.3%
#6

Meta AI

Muse Spark

69.7%
#7

Qwen

Qwen3.6 Plus

69.7%
#8

MiniMax

MiniMax M2.7

68.7%
#9

Grok

Grok 4.1 Fast (Reasoning)

68.0%
#10

Anthropic

Claude Opus 4.1

66.3%
#11

Google

Gemini 3 Flash

66.3%
#12

Google

Gemini 2.5 Pro

66.0%
#13

MiniMax

MiniMax M2.5

66.0%
#14

Anthropic

Claude Sonnet 4.5

65.7%
#15

Google

Gemini 3.1 Flash Lite

65.3%
#16

Moonshot AI

Kimi K2.5

65.3%
#17

DeepSeek

DeepSeek V3.2

65.0%
#18

Anthropic

Claude Sonnet 4

64.7%
#19

OpenAI

GPT-5

63.7%
#20

Z.ai

GLM-5

63.3%
#21

Google

Gemma 4 31B

62.0%
#22

Google

Gemini 2.5 Flash

61.7%
#23

OpenAI

GPT-5.4 Mini

61.3%
#24

OpenAI

GPT-4.1

61.0%
#25

Xiaomi

MiMo-V2-Pro

60.7%
#26

NVIDIA

Nemotron 3 Super

60.0%
#27

Anthropic

Claude Sonnet 4.6

58.7%
#28

Grok

Grok 4.20 (Reasoning)

58.0%
#29

Qwen

Qwen3.5 397B A17B

58.0%
#30

OpenAI

GPT-5.4 Nano

57.3%
#31

Google

Gemini 2.5 Flash Lite

48.0%
#32

Meta AI

Llama 4 Maverick

46.0%
#33

OpenAI

GPT OSS 120B

43.7%
#34

OpenAI

GPT-5 Nano

40.0%
#35

OpenAI

GPT-5 Mini

35.7%
#36

Anthropic

Claude Opus 4

33.7%
#37

Meta AI

Llama 4 Scout

25.8%
#38

Meituan

Longcat Flash Chat

25.7%
#39

Grok

Grok 4.1 Fast

22.0%
#40

Amazon

Nova 2 Lite

17.7%
#41

Grok

Grok 4.20

17.3%
#42

Baidu

ERNIE 5.0 Thinking

6.7%
#43

Baidu

ERNIE 4.5 300B A47B

2.3%