Kling 4.0 · Clipia

Kling 4.0 up to 30-second videos with sound

A clip of up to 30 seconds together with speech, ambience and music. Build the scene from text, image and video references or keyframes.

Create with Kling 4.0 Watch examples

per clip: up to 30 s
resolution: 720p / 1080p
in every generation: sound and speech

In short

What is Kling 4.0

Kling 4.0 is a Kuaishou video model that generates picture and sound in a single pass. Clips run from 3 to 30 seconds in 720p or 1080p. Pick a mode in the editor: from text, from references or from keyframes.

3–30 s: clip length
up to 10 images: references per request
up to 10 frames: scene keyframes
sound always on: speech, ambience and music

Text → video · speech

A lighthouse keeper in a storm

Six seconds: a keeper in a yellow raincoat turns from the window and calmly delivers his line. The voice, rain, thunder and surf are generated together with the picture while the camera slowly pushes in on his face.

How to set up the scene

Describe the place, the character and the action, put the line in quotes and set the delivery: "says in a calm low voice". List the ambient sounds separately — rain, distant thunder, surf.

Open this mode

Image references → video

A glass hummingbird lands on a bottle

Two images go in: a glass hummingbird and a perfume bottle on marble. In a five-second shot the bird comes to life, circles the bottle and lands on the cap; the bottle shape, the liquid colour and the bird's glass body are preserved. You can hear wingbeats and a soft chime.

How to set up the scene

Name the reference roles in the text: "the hummingbird from @image_1 lands on the bottle from @image_2". List what must stay exact — silhouette, material, colour — then add light and camera movement.

Open this mode

Keyframes → video

From a dark studio to a marble table

The clip starts on the glass hummingbird in a dark studio and ends on the perfume bottle on sunlit marble. The model builds the transition itself: the bird takes off, a flash of light dissolves the darkness and the camera arrives at the last frame.

How to set up the scene

Set the frames with @first_frame and @last_frame and describe what happens between them. For more control add intermediate frames — up to 10 in total.

Open this mode

Open an available mode of the model. Settings and cost are shown in the editor.

Create with Kling 4.0

Parameter	Value	Note
Duration	3–30 seconds	The editor offers 3, 5, 8, 10, 15, 20 and 30 seconds
Resolution	720p, 1080p	No 4K or HDR in this version
Modes	Text, references, keyframes	Three catalog entries: from text, from references and from keyframes
Sound	Native, always on	Speech, ambience and music are generated with the video
Aspect ratio	Auto, 16:9, 9:16, 1:1, 21:9	Auto fits the frame to the references
Inputs	Text up to 8,000 characters, images, videos, characters, voices	Up to 10 images, 5 videos, 7 characters and 3 voices, 15 in total

Text to video

A scene with lines, ambience and music from a single description — up to 30 seconds.

Open this mode

Keyframes

Set the first and last frame plus up to 8 in between, and the model guides the scene through them.

Open this mode

Image and video references

Up to 10 images and up to 5 videos: a character's look, a product, a location, camera movement.

Open this mode

Characters and voices

Saved characters and voices keep your hero recognisable from one clip to the next.

Open this mode

Parameter	Kling 4.0	Kling O3
Duration	Up to 30 seconds	Up to 15 seconds
References	Up to 10 images and 5 videos, characters and voices	Up to 7 images, or up to 4 with one video
Keyframes	Up to 10: first, intermediate, last	First and last frame

Choose the basis

Text only, a set of references or a chain of keyframes.

Link materials to the text

Refer to files with @-mentions: @image_1, @video_1, @first_frame.

Set length and sound

Pick 3 to 30 seconds and describe the lines, ambience and music.

Choose your materials and an available mode.

Create with Kling 4.0

A 5-second clip costs 75 credits in 720p and 110 credits in 1080p. Sound is included.

If a video reference is attached, the length of the source video is also counted.

Put lines in quotes

Write speech verbatim and say who speaks and in what voice. A long monologue works better as short phrases.

Call references by name

@image_1, @video_1, a character name — this tells the model which file does what.

Describe sound separately

List the background, effects and music in a separate sentence: sound is always on, and without a description the model picks it itself.

A talking character

A host or hero delivering a line with voice and background — no separate voice-over.

Product clip from photos

A product from image references in a new scene, with its shape and colour preserved.

Storyboarded scene

Keyframes set the story's route and the model fills in the motion between them.

Before a request, check the mode's availability in the catalog and get a cost estimate. Parameters depend on the mode.

Clipia documentation

FAQ

Is Kling 4.0 available in Clipia?

Yes. Available modes open from the generation buttons on this page; the editor shows the current choice.

How is Kling 4.0 different from Kling 3.0 and O3?

Kling 4.0 makes clips of up to 30 seconds with sound in every generation, accepts up to 10 image references, video references, characters and voices, and up to 10 keyframes. 4K and HDR, available in Kling 3.0 and O3, are not supported in this version.

How long can a clip be?

From 3 to 30 seconds. The editor offers 3, 5, 8, 10, 15, 20 and 30 seconds.

How much does a generation cost?

A 5-second clip costs 75 credits in 720p and 110 credits in 1080p; sound is included. The exact amount for your settings is shown in the editor before you start.

Are the demos on this page real generations?

Yes. All three clips were made with Kling 4.0 in 720p with sound and no editing: a scene with a spoken line from text, a clip from two image references and a transition between a first and a last frame.

Will Kling 4.0 be in the Clipia API?

Yes, for modes with active status in the Clipia catalog. Check parameters and cost before creating a task.

Create a video

Shoot your own scene in Kling 4.0

Describe the scene and the lines, add references or keyframes. The cost is shown before you start.

Create with Kling 4.0

Kling 4.0 demo

Watch the clip and explore its motion, detail, and camera work.

Watch demo

Open Kling 4.0 in the editor

Kling 4.0 — русская версия

Kling 4.0 · Clipia

Kling 4.0 up to 30-second videos with sound

A clip of up to 30 seconds together with speech, ambience and music. Build the scene from text, image and video references or keyframes.

Create with Kling 4.0 Watch examples

per clip: up to 30 s
resolution: 720p / 1080p
in every generation: sound and speech

In short