Cinematic prompting
Direct subjects, movement, dialogue, shot scale, camera motion and transitions in natural language.
Kling 4.0 video generation
Turn a detailed prompt into a cinematic video or animate a source image. Choose the duration and resolution, then follow the asynchronous result online.
Output
Your video will appear here
Choose text or image mode, add the prompt and submit a Kling task.
The model, explained
Kling 4.0 is the newest Kling video model for turning written direction or start and end frames into cinematic clips of up to 30 seconds. Kling 4.0 Flash and Kling 3.0 Turbo remain available for faster drafts.
Text-to-video gives you a blank canvas for subjects, action and camera language, while image-to-video animates a supplied frame and carries its visual identity into motion.
Generation runs asynchronously in FramePack. FramePack tracks the task and privately stores the finished video for preview and download.
Fast cinematic generation
Use a focused set of controls to move from a shot idea or still image to an HD video.
Direct subjects, movement, dialogue, shot scale, camera motion and transitions in natural language.
Upload a start frame, and with Kling 4.0 an optional end frame, then describe how the subject, environment and camera should move.
Choose 720p or 1080p, a whole-second duration that fits the selected model (up to 30 seconds with Kling 4.0), and the aspect ratio for text-to-video.
Follow the generation task while it renders, then preview or download the privately saved video.
Three simple steps
Pick a starting point, direct the motion, and let the provider render the finished clip.
Choose text-to-video for a new scene or image-to-video when an existing frame should anchor the result.
Describe action and camera behavior, add the source image when needed, and set duration, resolution and framing.
Track the asynchronous task, then preview and download the completed Kling 4.0 video.
A clear shot plan helps the model prioritize the movement that matters most.
Questions, answered
Yes. You can generate a new clip from a prompt, or upload a start frame and, with Kling 4.0, an optional end frame for motion guided by the prompt.
Kling 4.0 accepts an optional JPG, PNG or WebP start frame and end frame; Kling 4.0 Flash accepts a start frame only. Kling 3.0 Turbo image-to-video uses one JPG or PNG source image. Each image can be up to 10 MB.
Timing depends on duration, resolution and provider queue capacity. The page continues checking the asynchronous task while it renders.
When the generation is complete, FramePack saves the video privately and shows authenticated preview and download controls on the page.
Describe the scene or upload a frame, then generate an HD clip with Kling 4.0.
Create with Kling 4.0