Direct 3–30 second videos with text, keyframes, or named references. Describe motion, camera direction, dialogue, and sound.
No video input: 15.7 credits/second at 720p, 23.5 at 1080p. With video: 9.4 or 14.1 credits × (output seconds + max(output seconds, total input video seconds)). The request total rounds up. Video estimates are verified before charging.
Kling 4.0 AI Video Generator
Shape a complete shot with text, multiple keyframes, and multimodal references.
Create 3–30 second videos in 720p or 1080p. Start with a prompt, direct key visual moments, or bring characters, products, and voices together. Generation starts at 48 credits for 3 seconds at 720p without video input.
What is Kling 4.0?
Kling 4.0 is a video generation model for creative direction across text, images, video, and voice references. Use it for product storytelling, character scenes, social videos, and longer continuous shots. Flux Context provides a single generator with visible credit estimates and task recovery.
Kling 4.0 Core Capabilities
Three workflows for ideas, keyframe direction, and reference consistency.
Text-to-Video
Describe a scene, motion, camera movement, and dialogue.
- •3–30 second output
- •720p and 1080p
- •Widescreen, portrait, square, and 21:9
Multiple Keyframes
Direct the beginning, middle, and end of a shot.
- •First and optional last frame
- •Up to 8 ordered middle frames
- •First-frame aspect ratio
Omni Reference
Combine visual direction with named subjects and voices.
- •Up to 15 combined materials
- •Image and video subjects
- •Named and character-bound voices
Official Kling 4.0 Examples
Explore Kling's official demonstrations of realistic motion, visual references, and multi-keyframe storytelling.
Watch official Kling 4.0 examplesFrequently Asked Questions
What can I create with Kling 4.0?
Generate videos from text, animate a first frame with optional middle and last frames, or combine images, videos, named subjects, and voices in reference mode. Outputs run from 3 to 30 seconds at 720p or 1080p.
How many credits does Kling 4.0 use?
Without video input, 720p costs 15.7 credits per output second and 1080p costs 23.5. A 5-second video costs 79 or 118 credits after rounding up the request total. Images and voice references do not add billable seconds.
How do video references affect the price?
With any video reference, including a video subject, billable seconds equal output duration plus the larger of output duration and total input video duration. The rate is 9.4 credits at 720p or 14.1 at 1080p. Input durations keep their fractional seconds, and the final request total rounds up.
How many references can I use?
Reference mode supports up to 10 images, 5 videos, 7 named subjects (including at most 3 video subjects), and 3 standalone voices. These arrays share a limit of 15 materials. Reference videos and video subjects share a limit of 5 clips and 30 seconds.
How do multiple keyframes work?
Add a first frame, up to 8 middle frames in chronological order, and an optional last frame. Total frames cannot exceed 10 or output seconds plus one, whichever is smaller. The output follows the first frame's aspect ratio.
Can I recover a generation after leaving the page?
Accepted tasks remain in My Creations. Temporary network errors do not mean generation failed. Check the latest status without submitting a new task.
Create your next video with Kling 4.0
Choose your inputs, direct the scene, and review the credit estimate before generating.