A recent analysis of AI models on November 16, 2026, reveals that Korean models are leading in several benchmarks and download counts on Hugging Face. Out of 55 ranked boards on Hugging Face's official benchmarks, Korean models secured the top position on 14 boards. Bidraft models, in particular, achieved first place on 12 of these boards. In terms of downloads, POCKET-Darwin-180B-GGUF is the leading open LLM in Korea, with over 528,000 downloads in the last 30 days. Bidraft models also show strong performance, with four of their models appearing in the top 30 download list for Korea. AI
IMPACT Highlights the growing strength and popularity of Korean AI models in the open-source community.
RANK_REASON Analysis of benchmark and download data for AI models. [lever_c_demoted from research: ic=1 ai=1.0]
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →