The race for artificial intelligence (AI)-generated video has just begun, and China is poised to challenge OpenAI with Kling, a new AI video model that promises to surpass the performance of Sora, OpenAI’s recently much talked-about model.
Kling was developed by Kuaishou, a Chinese video platform, and seems to have all the makings of the new king of AI videos. In addition, the model is capable of generating high-definition (1080p) videos up to two minutes in length, with smooth movements and simulation of real-world physics. These capabilities make it a very competitive contender for Sora, which is currently undergoing testing and is not yet available to the public.
Kling uses advanced 3D face and body reconstruction technology, which enables it to create videos with realistic movements and complex facial expressions. In addition, the model can generate videos in different aspect ratios and show articulation movements and facial expressions due to its ability to train at different resolutions.
Kling’s availability is currently limited to Chinese users, but Kuaishou has opened a waiting list for those who wish to try the model. This approach could allow Kling to quickly gain a user base and gather valuable feedback while Sora remains locked in OpenAI labs.
The challenge between Kling and Sora is just beginning, but it is clear that both models have the potential to revolutionize the video industry. With new models entering the scene and research constantly moving forward, we can expect to see longer and longer, more realistic and impressive videos in the years to come.