Motif Technologies received the Venture Business Association Chairman's Award at the 4th Asia AI Awards ceremony held at the Korea Press Center in Jung-gu, Seoul, on November 5, recognized for demonstrating world-class performance by carrying out every stage from architecture design to pretraining using its own technology without relying on overseas models. The startup specializing in AI foundation models was established in February 2025 after spinning off from Moreh, an AI infrastructure company, and is developing its own multimodal models covering text, images, and video. Motif 3, a large language model with 314 billion parameters developed using the company's own technology, scored 47 on Artificial Analysis' Artificial Analysis Intelligence Index (AAII), ranking first among South Korean companies, ninth among LLMs globally, and third worldwide among open-weight models. The company completed the model in about five months with a core team of 30 and around 800 graphics processing units (GPUs). It has released major models as open source on Hugging Face and is participating in Microsoft Foundry and Marketplace and the NVIDIA Inception Program. CEO Lim Jeonghwan said it is possible to compete on equal footing on the global stage without relying on overseas foundation models.