Kuaishou Updates Kling AI With 30-Second Native 4K Video Generation Capabilities

Structured Digest of the Announcement

Kuaishou has released a preview of Kling 4.0, a video generation model that introduces extended clip durations and enhanced reference management systems. This update focuses on improving spatial continuity and synchronization between audio and visual elements during the generation process. By providing users with tools for character consistency and native high-resolution output, the platform aims to address technical limitations previously observed in generative video workflows. The preview is currently accessible to creators via the official Kling AI web portal.

Key Digest

10 Essential Facts at a Glance

Kling 4.0 now supports native generation of video clips up to 30 seconds in length.
An experimental Long Video mode is available for producing continuous 120-second sequences at 1080p.
Generated clips can be extended to 60 seconds after the initial creation process.
The Omni Reference system allows for up to 15 reference elements from a pool of 50 uploaded files.
Users can reference up to 30 images, 10 video clips, and 10 audio clips per generation.
The model supports multi-shot generation designed to maintain spatial and character continuity.
Audio generation, including dialogue and ambient sound, occurs simultaneously with video generation.
Lip movement is automatically synced to spoken dialogue across multiple languages and accents.
The update introduces native 4K output at a resolution of 3840×2160.
The Kling 4.0 preview is accessible through the dedicated Kling4.org web platform.

What This Means

The integration of the Omni Reference system represents a shift in how generative video models manage visual consistency across multiple shots. By allowing users to pull from a pool of images, video clips, and audio files, the model attempts to maintain character identity and environmental details without the need for repetitive prompting. This approach addresses common industry challenges regarding scene resets and visual drift, which often occur when camera angles or perspectives change in AI-generated sequences. The ability to lock specific attributes, such as lighting or costume, directly into the generation pipeline provides a more structured workflow for narrative production.

Simultaneously, the move toward native 4K output and integrated multi-language audio generation signifies a change in production requirements. By shifting these tasks from post-production software into the initial model generation phase, the tool reduces the number of external steps required for content creation. The inclusion of lip-syncing for multiple languages and ambient sound generation within a single pass suggests a focus on streamlining the production of localized media. As these models move toward longer durations and higher resolutions, the requirement for manual intervention in editing and dubbing processes may decrease, altering how small teams and individual creators approach the development of complex video projects.

Questions Worth Watching

How will the integration of multi-shot continuity affect the average time required to produce short-form narrative content?
To what extent will native 4K generation reduce the reliance on secondary upscaling tools in professional media workflows?
What challenges remain in maintaining character consistency when moving beyond the current 15-reference element limit?

Loading briefing details...