Kling AI Video Generator Combines Text and Images for Cinematic Quality
Kuaishou Technology's Kling AI, launched in June 2024, offers advanced video generation. Its latest version, Kling 2.6, enables simultaneous audio-visual generation, streamlining content creation.
Chinese technology company Kuaishou Technology has developed an AI video generation platform named Kling AI. Launched in June 2024, the platform's latest iteration, Kling 2.6, introduces simultaneous audio and video generation, significantly enhancing the workflow for social content creators.
Kling AI transforms text descriptions and static images into high-quality cinematic videos. It utilizes a Diffusion-Convolutional Neural Network architecture combined with a 3D Spatiotemporal Joint Attention Mechanism, allowing it to understand motion across space and time.
A newer feature, Kling Video O1, employs Chain of Thought reasoning to 'understand' prompt physics and logic before rendering. It also supports text-based editing, enabling users to swap objects or alter backgrounds without reshooting.
Users can upload still images and animate them with surprisingly natural motion. The Elements system further allows for combining up to four reference images to maintain character consistency across generated videos. Kling AI's free tier offers limited credits, while paid plans start around $10 per month, providing better quality and more features.