EXAONE 3.0 7.8B Instruction Tuned Language Model

TL;DR

EXAONE 3.0 7.8B is LG AI Research’s first open LLM in the EXAONE family, released for non-commercial and research use to accelerate open AI development. The model uses a decoder-only transformer with RoPE and GQA, a bilingual BBPE tokenizer (102,400 vocab), and a two-stage post-training regime (SFT + DPO) to enhance instruction-following. Across English and Korean benchmarks, EXAONE 3.0 demonstrates competitive English performance and superior Korean capabilities, supported by extensive real-world, math, coding, reasoning, and general evaluations, as well as robust open-leaderboard results. Responsible AI practices include data compliance, red-teaming, and explicit licensing constraints to mitigate misuse while enabling broad research and enterprise exploration in a controlled open framework.

Abstract

We introduce EXAONE 3.0 instruction-tuned language model, the first open model in the family of Large Language Models (LLMs) developed by LG AI Research. Among different model sizes, we publicly release the 7.8B instruction-tuned model to promote open research and innovations. Through extensive evaluations across a wide range of public and in-house benchmarks, EXAONE 3.0 demonstrates highly competitive real-world performance with instruction-following capability against other state-of-the-art open models of similar size. Our comparative analysis shows that EXAONE 3.0 excels particularly in Korean, while achieving compelling performance across general tasks and complex reasoning. With its strong real-world effectiveness and bilingual proficiency, we hope that EXAONE keeps contributing to advancements in Expert AI. Our EXAONE 3.0 instruction-tuned model is available at https://huggingface.co/LGAI-EXAONE/EXAONE-3.0-7.8B-Instruct.