Kuaishou's Kling 3.0 on Pixel Dojo
Kling 3.0 AI Video Generator
Kling 3.0 makes cinematic 1080p video with its own sound, from a motorcycle throwing spray through a neon tunnel to a lightkeeper speaking his line over the storm.
Kling 3.0 is Kuaishou's Kling AI video model for text to video and image to video. On Pixel Dojo it makes clips of 3 to 15 s with native sound and dialogue, plus a multi-shot mode that strings up to 6 shots into one clip. We made the Kling 3.0 videos on this page on Pixel Dojo, and each one comes with its exact prompt, so you can start from ours and make it your own.
27.5 credits ($1.72) per 5 s clip with audio. Included in every plan, from $10/month.
Already a member? Sign in
By Brian Lovett · Updated
What you can create with Kling 3.0
Four scenes we made with Kling 3.0: high speed action, spoken dialogue, a sweeping aerial and a claymation short. The first three are Pro clips at 1920 by 1080 with native sound, and the claymation is Standard at 720p. Tap Use this prompt on any clip to start from our wording, then set Tier and Duration to match the clip you like.
High speed action with real physics. The rider leans hard through the flooded tunnel, a rooster tail of spray kicks off the rear tire, and she powers away toward the far end under cyan and magenta light. Start from this prompt for your own chase scene.
Low tracking shot racing alongside a motorcycle rider in a cherry red leather suit and mirrored helmet as she leans hard through a flooded neon underpass at night, the rear tire carving through ankle deep water and throwing a towering rooster tail of spray lit magenta and cyan, then she snaps upright and rockets toward the tunnel exit as the camera whips past her, reflections streaking across the wet concrete, anamorphic flares, 35mm film grain, teal and crimson grade, no license plate, no logos. Sound: engine roar echoing off the tunnel walls, a rising gear shift whine, hiss of water spray, no music.
Dialogue with native sound. The lightkeeper lifts his eyes and speaks his line with his mouth in sync, over rain and distant thunder, lit warm by the lantern against a cold blue storm. Put the line in quotes and Kling 3.0 voices it.
Slow push in to a close-up of a weathered lighthouse keeper in his sixties with a grey beard and a yellow oilskin coat as he lifts his eyes from a brass storm lantern and speaks. [Lighthouse keeper, low gravelly voice]: "The light stays on tonight." Rain streams down the window behind him, warm lantern glow against cold blue storm light, 85mm shallow depth of field, fine film grain. Sound: rain on glass, distant thunder, wind moaning. No background music, no score, no soundtrack. No subtitles, no captions, no on-screen text.
Aerial scale in one move. The camera dives down the face of a towering waterfall through golden sun shafts, then pulls up to skim the churning river below. A strong opener for travel and nature edits.
FPV drone dive straight down the face of a colossal waterfall in a misty jungle canyon at sunrise, then pulling up to skim low over the churning river as birds scatter from the cliffs, golden light shafts through the spray, rainbows in the mist, ultra wide lens, crisp detail. Sound: thundering water, wind rush, distant bird calls. No music.
Claymation charm on the Standard tier. A clay hedgehog chef flips a pancake high out of a cast iron pan, lands it right back in and finishes with a big cartoon grin. Start from this prompt for playful animated shorts.
Stop motion claymation, medium shot with a gentle push in: a round clay hedgehog chef in a tiny white hat flips a pancake high above a cast iron pan in a cozy miniature kitchen, the pancake tumbling through the air and landing with a soft flop, visible thumbprint texture on the clay, warm practical lamp light, shallow depth of field. Sound: sizzling butter, a cheerful little hum, a wooden spoon tapping.
Where Kling 3.0 shines: motion, sound and scale
Kling 3.0 is built for shots that move. Give it a physical beat, like a tire carving through standing water or a drone dive down a waterfall, and it renders the spray, the lean and the camera move in one continuous shot. Our motorcycle clip runs 10 s at 1920 by 1080 and follows the rider from the first hard lean all the way down the tunnel.
Sound comes with the picture. Every Pro clip we made carries a stereo track generated from the same prompt, so the rain and thunder behind our lightkeeper, and the line he speaks, arrive already in sync. Write the dialogue in quotes with the speaker in brackets and Kling 3.0 voices it with matching lip movement.
Two tiers share one prompt box. Pro renders 1920 by 1080 at 16:9, 1440 by 1440 square and 1080 by 1920 vertical, made for ads, trailers and anything you publish. Standard renders 1280 by 720, 960 by 960 and 720 by 1280 at a lower credit rate, which suits social cuts and exploring lots of ideas. Both run any length from 3 to 15 s at 24 fps, both turn a start image into video, and multi-shot strings up to 6 shots into one clip.
How to get results like these from Kling 3.0
Every prompt on this page follows the same shape, and it works on both tiers. These two clips are Standard, 3 s at 720p with sound off, for 9 credits each. The tips below are how we write for Kling 3.0.
Standard at 720p: the red suited rider leans a blue sport bike through the neon underpass, with cyan and magenta light streaking across the wet concrete.
Low tracking shot racing alongside a motorcycle rider in a cherry red leather suit and mirrored helmet as she leans hard through a flooded neon underpass at night, the rear tire throwing a glittering fan of spray lit magenta and cyan, reflections streaking across wet concrete, anamorphic flares, 35mm film grain, teal and crimson grade. Sound: engine roar echoing off the tunnel walls, hiss of water spray, no music.
The lighthouse prompt on Standard: a warm hurricane lamp against a rain speckled blue window and a close, expressive performance.
Slow push in to a close-up of a weathered lighthouse keeper in his sixties with a grey beard and a yellow oilskin coat as he lifts his eyes from a brass storm lantern and speaks. [Lighthouse keeper, low gravelly voice]: "The light stays on tonight." Rain streams down the window behind him, warm lantern glow against cold blue storm light, 85mm shallow depth of field, fine film grain. Sound: rain on glass, distant thunder, wind moaning. No background music, no score, no soundtrack. No subtitles, no captions, no on-screen text.
- Lead with the camera. Every prompt here opens with the shot size and the move, like a low tracking shot, a slow push in or an FPV drone dive, then the subject, one action beat, the light and the lens.
- Name the physics you want to see. A towering rooster tail of spray, a pancake tumbling through the air, a dive down the face of a waterfall: Kling 3.0 gives motion to the details you write.
- End with a Sound line. List what you want to hear, such as rain on glass, distant thunder or an engine roar, and add no music when you want a clean effects track.
- For dialogue, put the speaker in brackets and one short line in quotes, then finish with no subtitles, no captions, no on-screen text so the words are spoken rather than printed.
- Pick the tier by where the video goes: Pro for 1080p ads, trailers and hero shots at 5.5 credits a second with sound, Standard for 720p social cuts at 3 to 4 credits a second.
- Use any length from 3 to 15 s. Five seconds covers one beat; our 10 s motorcycle clip had room for the lean, the spray and the run down the tunnel.
Kling 3.0 vs Kling 2.6 Pro: one night market prompt
Both models live in the same Kling Video tool, so we gave them one prompt at 5 s, 16:9, with sound. Kling 3.0 frames the handheld medium shot as written, with the cook, his apron and the string lit market all in frame as the wok flares. Kling 2.6 Pro goes close and moody on a tall column of flame and sparks. Choose Kling 3.0 when the person and the place tell the story, or when you want any length from 3 to 15 s; choose 2.6 Pro for a dramatic fire insert.
Same prompt: Handheld medium shot at a night market stall: a cook in his forties wearing a black apron tosses noodles in a steel wok as a burst of flame leaps up and lights his face orange, steam and sparks rising into the dark, string light bokeh behind him, 35mm lens, warm tungsten grade. Sound: roaring flame, sizzling oil, a clanging ladle, crowd chatter. No music.
Kling 3.0 Pro: a handheld medium shot with the cook, his apron and the market lights all in frame as the wok flares.
Handheld medium shot at a night market stall: a cook in his forties wearing a black apron tosses noodles in a steel wok as a burst of flame leaps up and lights his face orange, steam and sparks rising into the dark, string light bokeh behind him, 35mm lens, warm tungsten grade. Sound: roaring flame, sizzling oil, a clanging ladle, crowd chatter. No music.
Kling 2.6 Pro: a close, moody angle on the wok with a tall column of flame and sparks.
Handheld medium shot at a night market stall: a cook in his forties wearing a black apron tosses noodles in a steel wok as a burst of flame leaps up and lights his face orange, steam and sparks rising into the dark, string light bokeh behind him, 35mm lens, warm tungsten grade. Sound: roaring flame, sizzling oil, a clanging ladle, crowd chatter. No music.
| Kling 3.0 | Kling 2.6 Pro | |
|---|---|---|
| Credits for this clip | 27.5 | 25 |
| Output | 1920 by 1080, 24 fps, stereo sound | 1920 by 1080, 24 fps, stereo sound |
| Framing | Handheld medium shot as written, the cook's face and the market in frame | Close on the wok and the cook's hands |
| Fire | Flames flare up from the wok through the clip | A tall column of flame and sparks |
| Best for | Story shots where the person and the setting matter | Dramatic food and fire inserts |
| Clip lengths | Any whole second from 3 to 15 s | 5 or 10 s |
Kling 3.0 pricing on Pixel Dojo
Kling 3.0 is included in every Pixel Dojo plan, from $10/month, and uses credits per second of video, so you only pay for the length you make.
Standard is 3 credits a second, or 4 with sound. Pro is 4 credits a second, or 5.5 with sound. Turn sound off to pay the base rate.
At the settings on this page: 27.5 credits for a 5 s Pro clip with sound, 55 credits for the 10 s Pro motorcycle clip, and 9 credits for a 3 s Standard clip with sound off. Kling 2.6 Pro, in the same tool, is 25 credits for a 5 s clip with sound.
Kling 3.0 questions
What is Kling 3.0?
Kling 3.0 is Kuaishou's Kling AI video model for text to video and image to video. It makes clips of 3 to 15 s with native sound in two tiers, Standard at 720p and Pro at 1080p. On Pixel Dojo it runs in the Kling Video tool, which also offers Kling 2.6 Pro and a Reference to Video mode, and through the Pixel Dojo API.
What is Kling 3.0 best at?
Cinematic motion with real physics, like spray off a motorcycle tire or a drone dive down a waterfall, plus native sound and spoken dialogue in the same clip. It also handles stylized looks such as claymation. Every example on this page shows its prompt so you can start from it.
How much does Kling 3.0 cost?
Kling 3.0 is included in every Pixel Dojo plan, from $10/month, and uses credits per second. A 5 s Pro clip with sound is 27.5 credits, a 10 s Pro clip with sound is 55 credits, and a 3 s Standard clip with sound off is 9 credits.
How long can a Kling 3.0 video be, and at what resolution?
Any whole number of seconds from 3 to 15 s, at 16:9, 9:16 or 1:1. Pro outputs 1920 by 1080 at 16:9, 1440 by 1440 square and 1080 by 1920 vertical; Standard outputs 1280 by 720, 960 by 960 and 720 by 1280. Both run at 24 fps.
Does Kling 3.0 make sound and dialogue?
Yes. Sound is on by default and comes out as a stereo track in sync with the picture. Put the spoken line in quotes, keep it short with one speaker, and end the prompt with no subtitles, no captions, no on-screen text so the line is voiced rather than printed. Our Kling 3 prompting guide has more tips.
Kling 3.0 Standard or Pro: which should I use?
Pro for anything you publish: ads, trailers and hero shots at 1080p, 5.5 credits a second with sound. Standard for 720p social cuts and exploring ideas, at 3 to 4 credits a second. Both offer 16:9, 9:16 and 1:1 and every length from 3 to 15 s.
Can Kling 3.0 do multi-shot and image to video?
Yes. Turn on Multi-shot in the Kling Video tool and write up to 6 shots, each with its own prompt, and Kling 3.0 renders them as one clip priced on its total length. For image to video, add a start image, plus an optional end frame to set where the shot lands. Both are available through the API as well.
Make your first Kling 3.0 clip
Write a prompt above or start from any example on this page. One subscription covers every model on Pixel Dojo.
Start with Kling 3.0