Byte will share AI model progress in text-to-image and video generation on a large scale in July.

AI video model field welcomes new trends, with ByteDance and Kuaishou competing intensively. According to the latest news, ByteDance is scheduled to share its latest progress in AI model technology in areas such as Wenshengtu and Sora-like new videos on July 19, especially in the innovation of long videos and high dynamic generation. Previously, ByteDance announced that it will hold an AI technology forum in Singapore, focusing on world model exploration, and revealed that its research scientists will explore continuous high dynamic long video generation solutions.

At the same time, ByteDance has raised AI large models to the highest priority within the group, and multiple teams such as Douyin and Jianying are accelerating the research and development of AI video model applications. It is expected that new achievements will be announced in the near future.

On the other hand, Kuaishou is also active in the field of AI video large models. Since the release and invitation test of the Ke Ling model in June, Kuaishou has further open-sourced the graphic and text-based Wenshengtu large model in July, and has freely opened the Ke Ling Web version. The user response has been enthusiastic, with a significant increase in the number of applicants and users who have opened accounts, and the number of generated videos has exceeded seven million.

This AI video large model competition led by the two major short video platform giants marks the rapid progress and fierce competition of Chinese AI technology in the field of video generation, indicating the new possibilities for future video content creation and dissemination.