3B (multimodal), 1.5B (text), 0.5B (text)
Apache 2.0
First open-source HyperCLOVA X models released for commercial use under Korea's sovereign AI ecosystem initiative; SEED 3B is multimodal (text+image) designed for Korean linguistic and cultural context understanding with visual data comprehension; outperforms competing models in image/video understanding within Korean contexts; trained on high-quality Korean-centric data with years of accumulated expertise; SEED 1.5B is instruction-following language model for translation (Korean/English/Japanese) and business communications; SEED 0.5B trained using pruning and knowledge distillation at 39× lower cost than Qwen2.5-0.5B while outperforming it on most benchmarks; exceptional Korean language performance across all sizes; enables high-performance conversational interfaces in resource-constrained environments (mobile apps, smart home devices, wearables); small yet powerful models designed for fine-tuning to specific business needs; represents strategic investment in building Korea's independent AI capabilities.