Kling AI drops version 4.0 with stereo audio, multilingual lip sync, and a lighter Flash variant
decoder ·
Kling AI has released version 4.0 of its video generation platform, introducing higher resolution capabilities and more precise creative controls. The update allows users to generate up to 30 seconds of video in a single pass and supports up to ten keyframe images for detailed scene management. An enhanced Omni Reference system now accepts multiple assets, including images, videos, and storyboards, to guide the generation process. A new Creative Recreation feature enables users to modify specific elements like characters or backgrounds within existing video templates. The model also adds stereo audio and multilingual lip-syncing for over ten languages, while in-video text generation remains stable during camera movements. Additionally, a lighter Flash variant is available for faster, lower-cost production at 720p resolution. Although Kling 4.0 performs competitively against rivals like Wan 3.0 and Seedance 2.5, it shows no significant lead in areas such as physics and logical sequencing.
decoder