LG ships K-EXAONE 2.0 — Korea's largest open model, 750B parameters under Apache 2.0

LG AI Research published K-EXAONE 2.0 on Hugging Face on July 31, 2026: a 750B-parameter hybrid-attention mixture-of-experts model with roughly 37B active parameters per token (256 experts, 8 activated), a 262,144-token context window and 10-language coverage, built under South Korea's sovereign-AI foundation-model program and licensed Apache 2.0 for commercial use. LG reports a 70.1 average across 24 benchmarks, up from 63.3 for K-EXAONE 1.0.

What it means Sovereign-AI programs are now shipping frontier-scale open weights under permissive licenses — the self-hostable ceiling keeps rising.