The ridge
A swordswoman on a misty mountain path. The camera closes from a wide shot to her hand without losing her face or her costume.
Kling's newest video model renders one continuous take of up to 30 seconds at 1080p, with sound. Steer it with a prompt, with first, middle and last frames, or with up to ten reference images.
One continuous shot from 3 to 30 seconds, with sound, at 720p or 1080p. Three ways in on the same key: describe it, pin it to key frames, or hand it the faces, products and places to keep.
One take. You choose the length, and the length is what you pay for.
The default, with 720p for drafts. Sound is generated with the picture.
First, middle and last frames, or reference images, in a single job.
Ultrawide, beside 16:9, 9:16 and 1:1, from the same model.
Camera, action, dialogue and sound in one prompt. Choose 21:9, 16:9, 9:16 or 1:1.
kling/kling-4.0-t2v
A first, a middle and a last frame, up to ten images, and the model travels between them. The frames set the shape.
kling/kling-4.0-i2v
Text, key frames and reference images together: up to ten images of the characters, products and places to keep consistent.
kling/kling-4.0-omni
Kling 4.0 footage with its own sound. Under each clip, what to watch for.
Large-scale motion and long camera moves that hold together from the wide shot to the close-up.
A swordswoman on a misty mountain path. The camera closes from a wide shot to her hand without losing her face or her costume.
A grand piano on the bow of a ship in heavy sea, from the air down to the player. Water, spray and wind stay physical through the move.
Effects-heavy fantasy: light, smoke and fabric move together while the character stays the same person.
Several key frames along one timeline: a creature drawn in light travels through the room from one pinned moment to the next.
Thirty seconds at 21:9: a jump from an open cargo door and a wingsuit flight through a desert canyon, in the cinema shape a trailer needs.
A long brief followed closely, or five words turned into a whole performance.
A detailed brief: two friends meet on a British street at dusk, handheld camera, film look, a line of English dialogue. The street, the timing and the hug land where the prompt puts them.
“An old man ranting about AI.”
Five words. The room, the gestures, the voice and the lip-sync come from the model.
Thirty seconds is long enough for a story beat, a product moment or a full social spot without cutting.
Reference images keep shape, materials and brand details steady across angles and light.
Up to 30 seconds in one take, with sound and dialogue, at 21:9 if it is going on a big screen.
Pin the first, middle and last frames and let the model fill the motion between them.
9:16, 1:1, 16:9 and 21:9 from the same prompt and the same key.
You set the length when you submit, so the price is known before the job starts. Aspect ratio and the way in (text, key frames or references) do not change it.
| Output | 720p | 1080p |
|---|---|---|
| Per second | $0.157 | $0.2355 |
| 10 seconds | $1.57 | $2.36 |
| 30 seconds | $4.71 | $7.07 |
Kling's own list price. Kling still marks the 4.0 rates as a preview. If Kling changes them, the price here follows.
Direct relationships with model providers. Our direct relationships with model providers help us secure better commercial terms for our clients, early access to new models, and access to models that are not publicly available.
On the invoice you already get. Same key, same monthly bill, split by project or team like the rest of the catalogue.
Kling 4.0 is in the catalogue as kling/kling-4.0-t2v, kling/kling-4.0-i2v and kling/kling-4.0-omni. With a key already, it is a new model name in your call. Without one, tell us what you want to run and we come back with a key, usually the same day.