Since the advent of OpenAI's ChatGPT at the end of 2022 and its public release, large language models (LLMs), a cutting-edge AI technology, have expanded from the exclusive research domain of tech giants to widespread attention and application exploration around the world. After a year of rapid development, GPT-based large model technology has become a key strategic force driving national technological and industrial upgrading, attracting international wide attention and competition. Currently, the field of large models is presenting unprecedented technological innovation vitality and a global competitive situation. As of February 2024, more than a hundred different pre-trained large models have emerged in China, with participants including top domestic academic research institutions and internet technology companies, all striving to seize the initiative in this technological revolution.
Against this backdrop, an in-depth analysis is conducted on the industrial chain layout, basic resources, market participants, industry practices, and application scenarios of Chinese large models. At the same time, a comprehensive model evaluation dimension is constructed, which is of great significance for promoting the standardized and secure development of China's large model industry. Relying on their team of over a hundred analysts, Frost & Sullivan and LeadLeo Research Institute adopt an equal-weight anonymous voting system to carefully select 15 representative large models for multi-dimensional comprehensive evaluation. This initiative aims to systematically sort out the competitive landscape in the field of Chinese large models, provide strong reference and guidance for industrial development, and thus play a positive role in promoting the overall development of the national industry.
Based on research and analysis of Chinese large models in 2024, Frost & Sullivan (Frost & Sullivan, abbreviated as 'Frost & Sullivan') in collaboration with LeadLeo Research Institute released the latest '2024 China Large Model Evaluation Market Research Report'.
