Mi:dm 2.0 Korea-centric Bilingual Language Models
About
We introduce Mi:dm 2.0, a bilingual large language model (LLM) specifically engineered to advance Korea-centric AI. This model goes beyond Korean text processing by integrating the values, reasoning patterns, and commonsense knowledge inherent to Korean society, enabling nuanced understanding of cultural contexts, emotional subtleties, and real-world scenarios to generate reliable and culturally appropriate responses. To address limitations of existing LLMs, often caused by insufficient or low-quality Korean data and lack of cultural alignment, Mi:dm 2.0 emphasizes robust data quality through a comprehensive pipeline that includes proprietary data cleansing, high-quality synthetic data generation, strategic data mixing with curriculum learning, and a custom Korean-optimized tokenizer to improve efficiency and coverage. To realize this vision, we offer two complementary configurations: Mi:dm 2.0 Base (11.5B parameters), built with a depth-up scaling strategy for general-purpose use, and Mi:dm 2.0 Mini (2.3B parameters), optimized for resource-constrained environments and specialized tasks. Mi:dm 2.0 achieves state-of-the-art performance on Korean-specific benchmarks, with top-tier zero-shot results on KMMLU and strong internal evaluation results across language, humanities, and social science tasks. The Mi:dm 2.0 lineup is released under the MIT license to support extensive research and commercial use. By offering accessible and high-performance Korea-centric LLMs, KT aims to accelerate AI adoption across Korean industries, public services, and education, strengthen the Korean AI developer community, and lay the groundwork for the broader vision of K-intelligence. Our models are available at https://huggingface.co/K-intelligence. For technical inquiries, please contact midm-llm@kt.com.
Key insights
What problem does Mi:dm 2.0 address?
Mi:dm 2.0 tackles existing LLMs lacking Korean cultural alignment by using a curated data pipeline, custom tokenizer, and Depth-up Scaling to build 11.5B Base and 2.3B Mini models.
How does it differ from prior work?
Unlike prior LLMs trained on insufficient Korean data, Mi:dm 2.0 uses a proprietary filtering pipeline and synthetic data. Base outperformed Llama-3.1-8B on KoBBQ by 32.12%p overall.
What are the practical uses?
Released under MIT license for research and commercial use, the models target Korean industries, public services, and education. Mini is optimized for on-device deployment in constrained environments.
What limitations remain?
The authors note risks of generating incorrect, biased, or inconsistent responses. The model only supports Korean and English, and political orientation and ethical judgments need improvement.