Chinese company Kuaishou unveils advanced AI video production model, Kling

(MENAFN) China is making significant strides in developing AI models, with Kuaishou Technology's Kling being a prime example of this rapid progress. Earlier this year, OpenAI launched its Sora video production model, capable of creating one-minute high-definition videos. Although Sora is not yet widely accessible, another text-to-video model, Kling, has started to gain attention.

Kuaishou Technology, a well-known Chinese short video platform, introduced Kling, a model that converts text to video. Unlike Sora, Kling is available to users through a waiting list. Utilizing similar technology to Sora, Kling can produce 1080p HD videos lasting up to two minutes. The model can generate realistic, large-scale movements that accurately simulate the physical world, thanks to its foundation on the Diffusion Transformer architecture, which translates text prompts into dynamic images and videos.

Kling employs advanced 3D facial and body reconstruction powered by Kuaishou's 3D VAE technology, enabling the creation of videos in various aspect ratios. A demo video released by the company showcased Kling's capability to produce videos up to two minutes long at 30 frames per second. In contrast, Sora is currently limited to one-minute videos. Sample videos posted on social media platform X highlight Kling's impressive ability to replicate real-world physical properties accurately.

It's worth noting that Kling is not Kuaishou's only text-to-video model. In April, the company launched Vidu AI, a model capable of generating 16-second videos at 1080p resolution. This demonstrates China's ongoing efforts and rapid advancement in the field of AI-driven video production.

MENAFN10062024000045015839ID1108313716