Kling 4.0 and Kling 4.0 Flash released with new video creation capabilities
Kling AI has released Kling 4.0 with generation of up to 30 seconds of video and support for up to ten input keyframes. Access through MCP and CLI has been added. The Kling 4.0 Flash variant is limited to 720p and 20 seconds; the company is still developing API support.
Kling AI has released Kling 4.0, a model that can create up to 30 seconds of video at a time. A prompt can include up to ten keyframe images to define transitions between scenes, character states and how the story unfolds. According to the company, the Omni Reference system accepts up to 15 reference inputs in a single generation, including images and videos.
The Creative Recreation feature allows an existing video to be used as a template and products or characters to be replaced. According to the company, elements such as facial expressions, movements or backgrounds can be edited independently without changing the rest of the shot. The company also reports the addition of two-channel stereo audio and improved lip sync in more than ten languages. 10-bit HDR output at 4K and 1080p, along with video extension to up to two minutes, remain announced features that are still in development.
The Kling 4.0 Flash variant was released at the same time. According to the company, it is intended to generate video faster and at a lower cost, but its limits are 720p and 20 seconds. Kling 4.0 is available through MCP and CLI, enabling integration into automated workflows with AI agents. API support is still in development, and pricing for its use has not yet been announced.
Why it matters
Creators can specify how a video unfolds more precisely using keyframes and use existing footage as templates. When choosing the Kling 4.0 Flash variant, they must account for its lower maximum resolution and shorter duration. Access through MCP and CLI opens up video creation automation, while planning for API use is constrained by pricing that has yet to be announced and support that is still in development.
Release card
Kling 4.0
Kling AI
- Context
- The source reports expanded prompt understanding to 8 000 tokens.
- Inputs
- The model accepts text prompts, images and reference videos. It generates video with two-channel stereo audio.
- Availability
- The model is available in the video creation interface and through MCP and CLI. API support is still in development.
- Enables the creation of videos up to 30 seconds long with up to ten keyframes.
- Enables products or characters to be replaced in an existing video.
- Supports automated workflows with AI agents through MCP and CLI.
- It cannot generate more than 30 seconds of video in a single step.
- 10-bit HDR output at 4K and 1080p has only been announced.
- Video extension to up to two minutes is still in development.
The article compares Kling 4.0 with Wan 3.0 and Seedance 2.5. In particular, it does not credit the model with a lead in logical continuity between shots, physics or natural dialogue.
The card summarizes information from the article and any dated corrections, with a link to the original source. It is not our assessment of the model. It does not yet have a dedicated editorial profile. Model selection and other announcements →
Two audiences, two different impacts
What this means
For individuals
Video creation now allows up to ten keyframes as input, providing more specific control over scene changes and how the story unfolds.
For a business
Teams can incorporate video generation into automated workflows through MCP and CLI. However, API deployment still lacks announced pricing and available support.
ProcessesCheck the original
Event sources
only one source so far · 1 publisher, 1 independent. We count feeds from the same owner only once.