Skip to main content

Kling 3.0 Motion Control copies the moves, expressions and gestures of a 3 to 30 second driving clip onto one character photo, at 720p or 1080p. The result runs exactly as long as the driving video.

From $0.101/s20% below the publisher's list price

Workbench

Input

Character animation
108/2500
Optional. Use it only for what the driving clip does not show - the setting, the look, anything to add. at most 2500 characters
0/2500
Things to keep out of the clip - objects, looks or motion you do not want - up to 2,500 characters. Optional; it does not change the price. at most 2500 characters
1/1

Drop files or media links, paste, or browse · 1 file · Max 10 MB · JPG · PNG · WEBP · GIF

  • 1
1 / 1 files
The image_url supplied to Kling 3.0 fixes the character’s appearance; use a clear image with enough room for the requested movement, and let the prompt describe the action.
1/1

Drop a video here, or browse · 1 file · Max 90 MB · MP4 · WEBM

  • 1
1 / 1 files
For Kling 3.0 Motion Control, video_url supplies the performance to transfer: use a 3–30 second driving clip and check that both inputs meet the stated image-size requirements.
affects price
720p or 1080p. Higher tier costs more per second.
Keep original audio
Keeps the driving clip's own sound in the result. On by default. It does not change the price.
Character facing
Follow the videoFollow the image
Which way the character faces in the result: video (the default) follows the person in the driving clip, image follows the person in your photo. With image the driving clip can be at most 10 seconds; with video, up to 30. It does not change the price.
Shot structure
Custom shot listAutomatic
How the shots are planned: customize (the default) keeps to your inputs, intelligent lets the model plan the shots itself. It does not change the price.
Safety checkerOff

SpicyAPI does not run a safety checker on this model. Your prompt and media go to the model as written, and the model itself decides what comes back.

Checking your account before enabling submission and pricing.

Output

Ready for a new run
Example result

Missing the model you need?

KLING'S 4K FLAGSHIP

Kling 3.0 API at a glance

Set the first frame, pin the last one if you want it, switch sound on, and Kling 3.0 shoots it up to 4K.

Top resolution
4K
Draft at 720p, deliver the keeper at 4K.
Longest clip
15S
3 to 15 seconds on text and image to video.
Motion Control input
30S
Feed a 3-30 second driving clip; the result matches its length.
Sound, switchable
AUDIO
Turn native audio on or off per request.
Frame control
1ST+LAST
Set an opening frame and, optionally, pin the closing one.
Content modes
2
Single continuous shot, or let the model cut its own multi-shot edit.

Results

What you can make with Kling 3.0

  • First frame, last frame

    Pin where a clip starts and where it ends

    Give Kling 3.0 an opening frame, and optionally a closing one, and it fills in everything between. In our test the camera walked a woman down a fog-bound pier from a distant silhouette to a close, still frame that matched the closing image almost exactly.

    Image-to-video, 5 seconds at 720p: "one continuous shot, no cuts" kept the dusk light and framing steady for the whole turn.

  • Native audio

    Dialogue and room tone in the same file

    Turn on generate_audio and the line you write in the prompt comes back spoken, with the room's own sound underneath it. A rooftop toast, a crowd, clinking bottles - all one clip, no separate audio pass.

    Text-to-video, 6 seconds at 720p with sound: a handheld shot through the crowd, ending on a shared toast.

  • Motion Control

    Put someone else's moves on your character

    Upload one clear photo of a character and a 3-30 second driving clip, and Kling 3.0 copies the moves, the expressions and the hand gestures onto your character while keeping their face, outfit and setting. The result is exactly as long as the driving clip.

    Motion Control, 5 seconds at 720p: a driving clip's choreography copied onto this character, keeping her own face, suit and studio.

  • Up to 4K

    Draft cheap, render the keeper big

    720p and 1080p cost less per second; 4K is there once a shot is worth the render. Resolution and audio both change the price, so drafting small and finishing large is the normal workflow.

Content test

Is Kling 3.0 uncensored? What we tested

Content levels tested on Kling 3.0, by mode
LevelText to videoImage to videoCharacter animation
Mild
Not tested yet
Not tested yet
Not tested yet
Moderate
RefusedTested Sep 28, 2026
RefusedTested Sep 28, 2026
Not tested yet
Strong
RefusedTested Sep 28, 2026
RefusedTested Sep 28, 2026
RefusedTested Sep 28, 2026
Very strong
RefusedTested Sep 28, 2026
RefusedTested Sep 28, 2026
Not tested yet
Maximum
RefusedTested Sep 28, 2026
Not tested yet
Not tested yet
How we test

How each mode behaves

Text to video
shot_type single keeps the model on one setup most of the time, but it can still cut to a second angle on its own - a party scene and two of our other single-shot tests each picked up an extra cut, once with a lighting change across it. If the shot must not cut, add "one continuous shot, no cuts" to the prompt; that line held the shot together and kept the light constant in our retest of the same brief.
Image to video
The clip opens on your image and can end on a second one if you set last_image_url. Across 15 seconds a continuity slip is possible in the background - in one of our tests a truck's position and the building behind it did not match between the opening and a later angle in the same clip.
Character animation
Motion Control needs one clear, full or half-body photo of the character and a driving clip at least about 720 pixels wide or tall on each side. The result keeps the character's face, outfit and setting from the photo and runs exactly as long as the driving clip, from 3 to 30 seconds; anything shorter or narrower than that is rejected before the job starts.
All modes
Neither 4K output nor shot-list control makes Kling 3.0 unrestricted: Kling reviews the prompt and result, and rejected adult NSFW requests do not consume a generation payment. SpicyAPI adds no separate filter to Kling 3.0, and cannot disable Kling’s checks. MiniMax H3 Spicy is a separate unrestricted choice from Kling 3.0, with its own input options.

Use cases

What to build with the Kling 3.0 API

  • A dialogue beat with real sound

    Write the line, turn generate_audio on, and get a clip where the words, the room tone and the picture all arrive together - a toast, a confession on a piano bench, a goodbye at a train door.

  • Continue a scene from your own frame

    Take a still you already have - a product shot, a portrait, a frame from an earlier clip - and describe only what happens next. The look, the wardrobe and the setting come from your image, not from the prompt.

  • Reuse a performance on a different character

    Shoot or license one driving clip of a dance, a walk or a gesture, then run it through Motion Control on as many character photos as you need. The choreography stays identical; only the person changes.

  • A finishing render for a shot you already drafted

    Iterate the framing and timing at 720p, then send the same prompt at 4K once the shot works, for a hero clip, a product close-up or a title sequence.

Prompts

Kling 3.0 prompt examples you can copy

  • Image to video

    Lock the shot with one line

    Result from this prompt

    One continuous shot with no cuts, the golden-pink dusk light staying the same throughout. He turns from the rail toward the camera, pushes his wet hair back with one hand and smiles, then says, "Jump in, the water's perfect." The yacht rocks gently and the sail snaps in the evening breeze. Slow handheld push-in. Sound: waves slapping the hull, wind in the rigging, gulls far off.

    Instead of “He turns from the rail toward the camera, pushes his wet hair back with one hand and smiles, then says, "Jump in, the water's perfect."”

    Why it works. Without the opening line, our test cut to a second angle partway through and the light shifted from dusk to daylight. Adding "one continuous shot with no cuts" plus the light description held both the framing and the color steady end to end.

    Try this prompt
  • Text to video

    Dialogue with the sound written in

    Result from this prompt

    Night on the marble terrace of a Venetian palazzo during a gala. A woman in her early thirties in a backless emerald silk gown walks away from the camera toward the balustrade, candlelight catching the line of her back, the Grand Canal glittering below. She stops, looks back over her shoulder at the camera and says with a faint smile, "Don't wait up." Slow tracking shot behind her, shallow depth of field. Sound: a string quartet playing inside, the murmur of guests, water lapping against stone.

    Why it works. The line in quotes came back as spoken dialogue in our test, word for word, because the soundtrack was turned on and the sound cues were written into the same prompt as the shot.

    Try this prompt
  • Character animation

    Move a performance onto a new character

    Result from this prompt

    She performs the routine in the same warm-grey photo studio, the satin suit catching the light as she moves.

    Why it works. The driving clip supplies every beat of the choreography; the prompt only needs to say what the character wears and where she stands, since Motion Control already knows the moves.

    Try this prompt
  • Text to video

    A crowd scene with one clear line to hear

    Result from this prompt

    A summer rooftop party in Brooklyn at blue hour, string lights and the Manhattan skyline behind. A mixed group of friends in their late twenties and thirties dance and laugh: a woman with long braids spins under a man's arm, two men clink beer bottles, someone throws a handful of confetti. One continuous handheld shot weaving through the dancers, ending as the whole group raises their drinks and shouts together, "To summer!" Sound: upbeat funk music, chatter, laughter, bottles clinking.

    Why it works. Naming one line for the whole group to shout, at the very end of the brief, gave the model a single moment to land the dialogue on instead of guessing when several people should speak at once.

    Try this prompt
  • Image to video

    Describe only what happens after the frame

    Result from this prompt

    Lightning flashes and lights the whole forecourt white for an instant. She pushes off the truck, pulls her jacket collar up against the rain and walks out from under the canopy into the downpour toward the camera, rain bouncing off the concrete around her boots. Sound: a sharp crack of thunder, heavy rain drumming on the canopy roof, her footsteps splashing, the buzz of the fluorescent lights.

    Why it works. The opening frame already sets the woman, the truck and the gas station; the prompt never redescribes them and only adds the lightning, the motion and the sound that come after.

    Try this prompt

Compare

Kling 3.0 vs similar models: which one to pick

ModelModesFromContentPick it when
Kling 3.0This pageCharacter animation · Image to video · Text to video$0.067 / output secondFiltered — This model refuses requests beyond its own limits outright. Measured, not assumed — models we have not measured carry no label at all.You want the flagship: first-and-last-frame control, a switchable audio track, resolution up to 4K, and Motion Control for copying a performance onto your own character.
Kling O3Image to video · Reference to video · Text to video · Video editing$0.067 / output secondFiltered — This model refuses requests beyond its own limits outright. Measured, not assumed — models we have not measured carry no label at all.You need up to seven reference images fused into one shot, or you want to edit an existing clip instead of generating a new one.
Kling 3.0 TurboImage to video · Text to video$0.112 / output secondFiltered — This model refuses requests beyond its own limits outright. Measured, not assumed — models we have not measured carry no label at all.You want a text or image take at 720p or 1080p with sound always included, and do not need a shot list or Motion Control.
Kling 2.6Character animation · Image to video · Text to video$0.069 / output secondFiltered — This model refuses requests beyond its own limits outright. Measured, not assumed — models we have not measured carry no label at all.You are planning a fixed 1080p scene lasting 5 or 10 seconds, or a separate character animation guided by a 3–30 second performance clip.

Pricing

Kling 3.0 API pricing

A 5-second clip starts at $0.336 at 720p and $1.68 at 4K.

FAQ

Kling 3.0 API FAQ

About Kling 3.0

Kling 3.0 API combines first-and-last-frame direction, optional sound and shot lists for generation up to 4K; Motion Control remains a separate 720p or 1080p workflow. Kling 3.0 runs as an asynchronous task billed in US dollars per second of output; check the selected settings before submission, and retrieve the result by task ID or webhook. Neither 4K output nor shot-list control makes Kling 3.0 unrestricted: Kling reviews the prompt and result, and rejected adult NSFW requests do not consume a generation payment. SpicyAPI adds no separate filter to Kling 3.0, and cannot disable Kling’s checks. MiniMax H3 Spicy is a separate unrestricted choice from Kling 3.0, with its own input options. Before submitting to Kling 3.0, obtain the rights and consent for the source material and take responsibility for the resulting video.

Related models and next tasks