New Drug Development·AI Robots… LG 'Exaone Ecosystem' Outshines Meta for the First Time

[Financial News]  "Exaone 4.0 achieved first place in almost all categories, including mathematics, science, and agent tool usage areas." (Lee Hong-rak, co-research director of LG AI Research Institute)
LG AI Research Institute has unveiled its own AI model service, the 'Exaone Ecosystem', which has been developed with the AI technology accumulated over the past 5 years. LG AI Research Institute plans to supply the new Exaone model to companies other than LG to increase corporate market revenue. They are also accelerating the discovery of new businesses, such as implementing 'Physical AI' by equipping robots with Exaone.
LG AI Research Institute held the 'LG AI Talk Concert 2025' on the 22nd at LG Science Park in Magok, Gangseo-gu, Seoul, and showcased △ the first domestic hybrid AI model 'Exaone 4.0' △ precision medical AI model 'Exaone Path 2.0' △ enterprise AI agent 'Chat Exaone' △ AI platform 'Exaone Data Foundry' that produces high-quality data △ full-stack (integrated) solution 'Exaone On-Premise' that helps companies build enterprise agents in an independent environment from external sources, allowing them to use the Exaone model without security concerns. Both Lee Hong-rak and Im Woo-hyung, co-research directors, attended the talk concert.
The Exaone 4.0, which combines a large language model (LLM) and inference model, has achieved world-class performance. The multimodal AI model 'Exaone 4.0 Vision Language (VL)' surpasses the performance of Meta's AI 'Llama 4 Scout' model. Exaone commercialization is also active. The London Stock Exchange Group (LSEG) will commercialize a business intelligence service incorporating Exaone in the third and fourth quarters of this year. It predicts the direction of investment asset returns based on unstructured data such as LSEG data, news, and disclosures, and generates reports to assist decision-making for investors worldwide. From today, companies, public institutions, and research institutes can use the beta version of Chat Exaone, which was previously only used by LG employees.
Lee Hwa-young, head of AI Business Development at LG AI Research Institute, responded about the timing of Exaone revenue generation, saying "Although small, it is currently occurring." He added, "It is important to package and sell AI technology as B2B," and "LG CNS and LG Uplus, which are already in B2B business, have order amounts exceeding 100 billion won." 
LG AI Research Institute is also accelerating the discovery of new businesses using Exaone. Exaone Path 2.0 can reduce disease diagnosis time from the existing 2 weeks to within 1 minute. The head of the department said, "Exaone Path 2.0 instantly determines genetic mutations or overexpression and selects which new drug candidate is most suitable for the patient," adding, "We will create new businesses in the field of discovering new drug candidates in clinical trials in the bio industry, in collaboration with large pharmaceutical companies in Europe and the United States."
They also plan to make Physical AI, which equips Exaone on physical devices like robots and cars, a new revenue source. Im Woo-hyung, co-research director of LG AI Research Institute, said, "We will secure global competitiveness based on our own foundation model and apply it to various industrial sites beyond AI model development," adding, "We will create an AI ecosystem with global partners, possessing both versatility and expertise."

mkchang@fnnews.com Min-kwon Jang, Reporter