Kling 4.0 is now available on Artlist, with AI-generated video clips of up to 30 seconds, middle-frame controls and stereo audio. The new model from Kuaishou gives creators more control over how a scene develops, including the timing of a camera move or a change in the action.
The update also supports multiple image and video references, dialogue in nine languages and a 21:9 ultrawide format. It is part of the Artlist AI Toolkit, alongside the platform’s other video-generation models.
Table of contents
Kling 4.0 Video Length and Resolution
A single generation can run from 3 to 30 seconds, compared with the 15-second limit of Kling 3.0. The longer duration offers more room for a product demonstration, a short story or a social-media sequence within one continuous take.
Output is available at 720p or 1080p. Supported formats include 21:9, 16:9, 9:16 and 1:1, covering widescreen scenes, vertical videos and square posts. 4K and 10-bit HDR are not available at launch.
Middle Frames and Reference Inputs
Middle frames let creators guide points between the opening and closing of a clip. Start with an image, then place visual cues along the timeline to guide a reveal, movement or transition at a chosen moment.
Kling 4.0 accepts up to 10 images and five video references in a generation. These inputs can guide the appearance of a character, object or setting. The timeline demo illustrates the approach with a street chase that moves towards a ship above the city skyline.
Stereo Audio and Dialogue
The model generates stereo audio and lip-synced dialogue together with the video. Artlist lists support for dialogue in nine languages, with accents and dialects that include Indian English.
Audio can also be switched off for creators who prefer to add their own soundtrack later. This makes the feature useful for both dialogue-led clips and visuals intended for a separate voiceover or music track.
Kling 4.0 Flash and Artlist Availability
Kling 4.0 Flash is the lighter alternative. It supports clips from 3 to 20 seconds at 720p, with a start-frame input rather than middle- and end-frame controls. Voice references are not supported in Flash.
On Artlist, users can open the AI Toolkit, select video and choose Kling 4.0 or Kling 4.0 Flash. The full model is the more suitable option when a longer clip and control over intermediate moments matter; Flash offers a simpler route for shorter iterations.
For another creator-focused update, read about Adobe Firefly’s music, speech and sound-effect tools.