The AI video generation model Kling has taken the world by storm, offering an impressive 2-minute video generation capability at 30fps and 1080p quality. Developed by Kuaishou Technology, the maker of Kuaishou, a popular short video creation and viewing app in China, Kling is poised to revolutionize the AI filmmaking landscape. With its unique 3D Variational Autoencoder (VAE) for face and body reconstruction, Kling is capable of generating highly realistic videos that mimic the physical world. The model’s capabilities are already impressing filmmakers, who see it as a game-changer for AI-generated content.
What’s striking about Kling is its ability to generate videos that are on par with OpenAI’s still non-public, invitation-only Sora model. This has sparked excitement among creators, who are eager to get their hands on the technology. While access to Kling is currently limited to those with a Chinese phone number, workarounds have been shared by venture capital firm a16z partner Justine Moore and U.S. filmmaker Dustin Hollywood.
As Kling continues to gain traction, it’s likely to put pressure on U.S.-based AI video model providers to step up the quality of their generations and resolution. Whether they can match Kling’s capabilities in short order remains to be seen, but one thing is clear: the arrival of Kling is a significant development in the world of AI filmmaking.
