LG’s AI model EXAONE 4.0 ranks 11th globally

(LG AI Research)

LG Group’s artificial intelligence (AI) model has been recognized among the world’s best by a leading global evaluation agency.

LG AI Research announced Wednesday that its AI model EXAONE 4.0 ranked 11th globally—and No. 1 among South Korean models—in the latest Intelligence Index by global AI performance analysis firm Artificial Analysis.

EXAONE 4.0 earned a composite score in the 64-point range, tying for 11th place with Claude 4 Opus, the flagship model from U.S.-based Anthropic PBC.

Among open-weight models—those whose weights are publicly available for modification and redistribution—EXAONE 4.0 ranked fourth globally, following Alibaba’s Qwen3, DeepSeek R1, and Zhipu AI’s GLM-4.5.

In specific skill areas, EXAONE 4.0 ranked 7th in coding and 10th in mathematics, both within the global top 10. Based on corporate development performance, LG AI Research ranked 8th worldwide.

The institute noted that EXAONE 4.0 stands shoulder-to-shoulder with frontier models that are several to dozens of times larger in size, showcasing LG’s technological prowess on the global stage. It is also the first AI model in Korea to combine general-purpose and reasoning capabilities in a single hybrid system.

On July 15, LG AI Research released EXAONE 4.0 as an open-weight model on Hugging Face, making it available for research, academic, and educational purposes.

The model follows the March launch of EXAONE Deep, Korea’s first reasoning-focused AI. Both models have been listed in Epoch AI’s Notable AI Models, published by the U.S. nonprofit AI research organization.