EXAONE 4.0 32B

EXAONE 4.0 32B.

LG AI Research Released 15 Jul 2025 Family EXAONE The lab's model card

Benchmarks with a score

Ranked first, out of five or more models

In the top three

LG AI Research Jul 2025

Open weights Flagship

No Epoch Capabilities Index score yet

Where it ranks

Its best score on each benchmark, and where that ranks among every model with a score there. We rank each model once, at its best setting.

Benchmark	Score	Rank	Source
MMLU-Pro	81.8%	#58 of 151	Model card
Multi-IF	73.5%	#12 of 28	Model card
MMMLU	85.6%	#29 of 62	Model card
MMLU-Redux	92.3%	#18 of 38	Model card
KMMLU-Pro	67.7%	#4 of 7	Model card
RULER	88.2%	#22 of 38	Model card
IFEval	84.8%	#76 of 126	Model card
Ko-LongBench (LG)	76.9%	#3 of 3	Model card
LongBench	48.1%	#2 of 2	Model card
HELMET	58.3%	#1 of 1	Model card
KMMLU-Redux	72.7%	#1 of 1	Model card

Benchmark	Score	Rank	Source
GPQA Diamond	75.4%	#138 of 284	Model card
AIME 2025	85.3%	#71 of 143	Model card
HMMT 2025	72.9%	#42 of 49	Model card
HRM8K-KSM	87.6%	#2 of 2	Model card

Benchmark	Score		Rank	Source
LiveCodeBench v5	72.6%		#6 of 44	Model card
LiveCodeBench v6	66.7%		#57 of 76	Model card

Benchmark	Score		Rank	Source
Berkeley Function Calling Leaderboard (BFCL)	65.2%		#32 of 47	Model card
tau-bench	57.1%		#17 of 23	Model card

The benchmarks the lab chose to headline, with the scores it gave.

Every result in the longer documents the lab published with it. 1 of them isn't in the launch post.