The Hong Kong Generative AI Research Center (HKGAI) officially launched its latest customized local large model, HKGAI V3, and introduced Hong Kong's first productivity-level super agent at the "HKGAI V3 Large Model Launch and Ecosystem Collaboration Conference" held on June 3rd.

This platform has successfully broken through the current capabilities of agent technology, achieving a breakthrough performance of stable operation without intervention for up to 28 hours in testing. It can seamlessly complete multiple complex tasks such as data organization, reasoning analysis, report writing, and code development across different stages, marking the official entry of end-to-end full-process autonomous driving-level AI applications in Hong Kong into the productivity stage.

As a pioneer of the large model ecosystem in Hong Kong, the Hong Kong Generative AI Research Center is supported by the InnoHK Innovation Hong Kong R&D Platform, a key science and technology project of the Hong Kong Special Administrative Region government. Looking back at its technological evolution, the research center officially launched Hong Kong's first self-developed artificial intelligence large model, HKGAI V1, in February of last year. The iteration of the V3 version and the launch of the super agent demonstrate its rapid evolution from building basic model capabilities to empowering deep and complex application scenarios.

Industry experts pointed out that the release of HKGAI V3 and the super agent not only significantly enhances the automation ceiling of local complex business processes but also provides high-reliability localized technical support for the digital transformation of government and enterprise sectors in the Guangdong-Hong Kong-Macao Greater Bay Area. This agent technology, which balances long-term stable operation with multi-task collaboration, is becoming a core indicator for measuring the industrialization strength of next-generation AI large models.