Clevi independently develops ‘Ivy’, a platform powered by ultra-large generative AI
Clevi independently develops ‘Ivy’, a platform powered by ultra-large generative AI
Photo=Clevi
Clevi Corporation (Clevi) announced that it would launch the open beta service for ‘Ivy’, an ultra-large generative AI independently developed using LLM (large language model) technology.
Korean start-up Clevi has officially launched its proprietary ultra-large generative AI, ‘Ivy’, using LLM (large language model) technology. As a high-performance, low-cost model, Ivy delivers outstanding effectiveness and performance, with reasoning, specialised agent capabilities and a high instruction-following rate. Until now, South Korea had not achieved notable results in the AI market amid the dominance of US-based OpenAI and China’s DeepSeek. However, the official launch of the ultra-large generative AI ‘Ivy’ by Korean start-up Clevi has demonstrated that South Korea can also develop AI using its own technological capabilities.
Ivy provides powerful capabilities including △real-time learning and adaptation △multi-data processing △a code interpreter △expert model generation and △an agent support system, enabling it to be used across various fields, from individual users to enterprises.
Clevi CEO Lee Hwan-ho had been interested in AI and continued research and development even before establishing the company in 2022. Despite limited funding after founding the company, he successfully developed the proprietary model over a period of more than two years through an efficiently designed training pipeline, starting with pre-training. According to the company, the model development cost approximately 1.5 billion won, while operating costs average around 10 million won per month. This is expected to provide a breakthrough solution to the AI cost issues that companies are grappling with.
CEO Lee Hwan-ho emphasises that the 'Ivy' model delivers high efficiency and performance, enabling it to address the issue often cited as a drawback of generative AI: 'PoCs are easy, but completion is not.' This allows businesses to reduce labour costs and improve productivity by having AI perform repetitive and inefficient tasks. To minimise AI development costs, CLEVI introduced innovative technology that efficiently manages the development process by applying automated, container-based training node deployment technology. It is also noteworthy that costs have been significantly reduced through the efficient design of training pipelines and optimisation of communication networks between nodes.
The 'Ivy' model is characterised by the development of a Mixture of models for efficient model training and operation. It was developed by creating small models, distilling and combining them into a model, and incorporating attention and layer expansion to develop efficient models through improved combining and routing. Its support for a wide range of languages worldwide—not only English and Korean—through multilingual training and embedding optimisation is also noteworthy.
In particular, as Open AI in the United States and DeepSeek in China compete to advance their AI capabilities, the domestic large-scale generative AI model 'Ivy' is attracting industry attention after recording higher scores than the two existing models in various benchmark tests (△AIME 2024: 81% △MATH-500: 98.6 △SWE-BENCH: 51.1 △MMLU: 91.2).
In addition, as concerns about personal information leaks and security have grown with the increased use of AI Assistants, CLEVI has made the Ivy app available on a platform that allows users to register with only the minimum amount of personal information. Public institutions and businesses can also install it on-premises to suit their user environments, and the company is regarded as having addressed security concerns.
A CLEVI representative said, “The company will not stop at launching Ivy. In the future, it plans to unveil an agent that directly intervenes in, responds to and participates in conversations between individual and group users. In addition, CLEVI has completed R&D and is preparing to launch services for the AI-based active workflow agent “ARI” (Adaptive Reasoning Intelligence), which enables more efficient use of AI across various fields, including education and R&D as well as enterprise-wide enterprise resource planning (ERP) systems, and for methods of performing precise object recognition and real-time control in areas such as video content production and medical image analysis. We ask for your continued interest in CLEVI’s progress as it makes an impact in the global AI market through innovative technology, outstanding performance and cost savings.”
Meanwhile, Ivy (‘Ivy’) has been registered on the Google Play Store, and Ivy can also be downloaded and used from the company’s official website and the Ivy download page.
Reporter Seo Hui-won [email protected]
Source: Electronic Times (Reporter Seo Hui-won) | https://www.etnews.com/20250325000155