Kling 4.0 is here: 30-second video, stereo audio and a Flash version

Kling AI made version 4.0 available through its video generation service on 19 October 2026. The update includes longer clips from a single generation, keyframes for directing scene flow, multiple references and new audio features. The company also announced a separate version, Kling 4.0 Flash, for faster, lower-cost generation.

Longer generations and more reference options

Kling 4.0 can generate videos up to 30 seconds long in a single generation. Users can provide up to 10 keyframe images to guide scene transitions, character states or important story moments. According to the source, its prompt understanding capacity has also increased to 8,000 tokens.

According to Kling AI, Omni Reference accepts up to 15 reference items per generation. The system can use up to 10 images and 5 videos with a combined duration of no more than 30 seconds; the source also mentions support for up to 7 items. Storyboards, wireframe drawings, image grids, animatic videos and depth-map videos can also be used as references.

The Creative Recreation feature focuses on using an existing video as a template and replacing the products or characters in it. Kling AI says elements such as facial expressions, body movements, camera angles or backgrounds can be edited independently of the other parts of a clip.

Stereo audio, lip-sync and Flash

The update adds two-channel stereo audio and improved lip-sync in more than 10 languages. Kling AI’s examples include German, English, Japanese, Korean and Spanish, as well as different dialects and accents. The company also lists the ability to generate on-screen text in multiple languages and maintain its stability during camera movement.

Kling 4.0 Flash is aimed at faster, lower-cost generation. However, this variant has different limits from the main version: it can generate videos up to 720p resolution and 20 seconds long. The interface has also been updated. Kling 4.0 is said to be accessible via MCP and CLI, enabling automated workflows with agents.

Features not yet available and unanswered questions

10-bit HDR output at 4K and 1080p, along with a feature to extend videos to up to two minutes, are listed by Kling AI as “coming soon.” These options should therefore not be treated as currently available features. Kling 4.0 API pricing has not been announced; the source also says API support is coming soon.

For architecture, interior design and archviz teams, the main use case could be turning design narratives into short videos and directing specific scene elements with references. Support for inputs such as storyboards, wireframe drawings and animatic videos opens up the possibility of bringing visual flows from the design stage into video generation. However, the source does not mention a dedicated architecture tool or guarantee accuracy in measurements, geometry or materials. Outputs should therefore be treated as concept narratives or presentation drafts, not project documentation.

Teams building automated workflows can explore MCP and CLI access, but should account for the fact that API support and pricing have yet to be clarified. The source provides no information on hardware requirements, licensing terms or compatibility with existing projects. 4K HDR and two-minute video extension should not be treated as definite production features until they become available.