What can Kling 3.0 generate on Soup.video?

Kling 3.0 turns a text prompt or still images into an MP4 clip from 3 to 15 seconds. Skip images for text-to-video. Add a start frame (and optionally an end frame or character elements) for image-to-video. Multi-Scene builds up to 5 shots in one generation. Native audio is optional and includes lip-synced dialogue.

How do I use Kling 3.0?

Write a prompt in the Director bar. Leave images empty for text-to-video, or attach a start frame to animate a still. Choose Single or Multi-Scene, then set Standard / Pro / 4K, duration, aspect ratio, and audio. Cook. You can close the tab; Soup.video emails you when the clip is ready.

How does text-to-video work?

Do not attach a start frame, end frame, or character images. The prompt becomes the whole shot. Set ratio, mode, and duration, then cook.

How does image-to-video work?

Upload a starting image and describe the motion. You can also add an end image so the clip lands on a specific last frame. Aspect ratio follows the still.

How does multi-shot work?

Turn on Multi-Scene. Each scene has a prompt and a duration (minimum 1 second). You can define up to 5 scenes. The sum of shot lengths must equal the clip duration, between 3 and 15 seconds.

Does Kling 3.0 generate audio?

Yes. Turn audio on to generate dialogue with lip sync, sound effects, and ambient sound in the same pass. Audio works best in English and Chinese. On Pro, voice control is billed at a higher per-second rate.

What modes and prices does Kling 3.0 support?

Standard, Pro, and 4K, billed per second. Standard is ✦ 12/s without audio and ✦ 18/s with audio. Pro is ✦ 16/s, ✦ 25/s with audio, and ✦ 29/s with audio plus voice control. 4K is ✦ 60/s with or without audio. A 3-second Standard clip without audio is ✦ 36.

Kling 3.0 AI Video Generator

A massive alien spaceship, miles long and covered in glowing blue circuits, slowly rises from the East River in Manhattan. Water cascades off its hull as it ascends past the Brooklyn Bridge. Helicopters scatter, car alarms blare across the city. The camera captures this from a rooftop in DUMBO, showing the ship dwarfing the entire skyline as golden sunset light reflects off its surface. Thunderous rumbling, water crashing, distant sirens.

Bamse, the world's strongest bear, a cartoon brown bear wearing blue overalls and a blue knit cap, lifts an enormous boulder over his head and hurls it across a sunny Swedish meadow. Wildflowers scatter in the wind. His friends, a small rabbit and a turtle with a shell full of gadgets, cheer from behind a wooden fence. Bright, colorful 2D animation style with thick outlines. Cheerful music, a triumphant fanfare, birds chirping.

Motorcycle racing through traffic at dangerous speeds

Tight close-up portrait of a Gen Z fashion model, early twenties, dewy glass-skin complexion with a light dusting of faux freckles, glossy bitten-lip stain in a muted berry tone, fluffy laminated brows, and a single pearl-studded graphic liner flick in chrome silver across one eyelid. Her hair is slicked back into a sleek low bun with two soft face-framing tendrils, small silver butterfly clips catching the light. She wears oversized vintage-style chrome chandelier earrings and a sheer mesh high-neck top layered under a cropped leather moto jacket. The camera holds an extreme close-up on her face, then slowly arcs around her in a subtle 45-degree orbit as she tilts her chin down, cuts her eyes directly into the lens with a confident deadpan stare, and exhales softly. A gentle breeze lifts the loose strands of hair across her cheekbone. Lighting is soft neon-tinged — cool lavender key light from the left blending into a warm peach rim light from the right, creating a duotone gradient across her skin. Background is an out-of-focus wash of deep magenta and teal bokeh. Cinematic editorial fashion aesthetic, shallow depth of field, 85mm lens compression, subtle film grain, photorealistic, 24fps, reminiscent of a modern i-D Magazine or Vogue Beauty film.

Cinematic drone footage of a dragon made of living fire soaring over a frozen Nordic fjord at twilight. Its wings trail streams of flame that melt the ice below, creating swirling clouds of steam. The dragon banks sharply and breathes a column of fire across the mountainside, igniting the snow in brilliant orange and gold. Northern lights shimmer green and purple in the sky above. Thunderous wingbeats, crackling fire, hissing steam, a deep otherworldly growl.

Why Kling 3.0?

Text or image

Skip stills for text-to-video. Add a start frame — and optionally an end frame — for image-to-video.

Native audio

Dialogue with lip sync, sound effects, and ambient sound in one pass.

Multi-shot

Up to 5 scenes in one clip. Each shot has its own prompt and duration.

Standard, Pro, 4K

Pick the quality tier. 4K is billed at the same rate with or without audio.

Up to 15 sec

Clips from 3 to 15 seconds — longer than earlier Kling versions.

Pricing

Credits per second

Mode	No audio	Audio	Voice
Standard	✦ 12/s	✦ 18/s	—
Pro	✦ 16/s	✦ 25/s	✦ 29/s
4K	✦ 60/s	✦ 60/s	—

Voice is Pro-only (audio plus voice control). 4K is the same rate with or without audio. Text-to-video and image-to-video share these rates.

Kling Video 3.0

Generate cinematic videos up to 15 seconds long from text prompts or images. Kling Video 3.0 improves on previous versions with longer output, stronger consistency across shots, and native audio generation including lip-synced dialogue.

What it does

Kling Video 3.0 turns text descriptions or still images into video clips. The model generates videos between 3 and 15 seconds — a significant jump from the 10-second limit of earlier versions. It handles realistic scenes, stylized content, and complex multi-step actions within a single generation.

You can also generate native audio alongside the video, including dialogue with lip sync, sound effects, and ambient sound — all in one pass. Outputs are .mp4 in Standard, Pro, or 4K.

How to use it

Everything happens in the Director bar at the bottom of the page. The model supports two main input modes — the images you attach pick which one runs. There is no separate mode label on the form.

Text to video — describe what you want to see. Skip start/end frames and character images. The model generates visuals (and optionally audio) from your description.
Image to video — upload a starting image and describe the motion you want. You can also provide an end image to guide where the video should land.
Settings — Standard, Pro, or 4K; duration 3–15 seconds; ratio 16:9, 9:16, or 1:1 (ignored when a start image is set); audio on or off.
Cook — the price in the button is mode × duration (and audio). Closing the tab after cook starts is fine; we email you when the clip is ready.

Multi-shot mode

For videos with multiple scenes, switch the Director to Multi-Scene. Each shot has a prompt and a duration. You can define up to 5 shots, with a minimum of 1 second per shot. The total duration of all shots must equal the clip length (3–15 seconds).

[
  {"prompt": "A woman walks through a sunlit forest", "duration": 5},
  {"prompt": "She stops and looks up at the canopy", "duration": 3},
  {"prompt": "Sunlight breaks through the leaves", "duration": 2}
]

Writing effective prompts

Structure your prompts to cover:

Scene setting — where and when the action happens, lighting conditions.
Subject details — what characters or objects appear, how they look.
Motion — what happens, how things move, camera behavior.
Audio (if enabled) — dialogue in quotation marks, ambient sounds, sound effects.

Example: A chef in a busy kitchen plates a dish with careful precision, steam rising from the food. Camera slowly pushes in on the plate. Sizzling sounds, kitchen chatter in the background, the chef says "Perfect."

Parameters

prompt: Required for a single scene. Describe setting, subject, motion, camera, and (if audio is on) sound. For multi-shot, each scene has its own prompt instead.
images: None for text-to-video. One still is the start frame. A second still is the end frame. Aspect ratio is ignored when a start image is set.
elements: Optional character references. Mention them in the prompt with @name. Character refs run as image-to-video.
mode: Standard, Pro, or 4K.
duration: 3 to 15 seconds. In multi-shot, shot lengths must add up to this total.
aspect_ratio: 16:9, 9:16, or 1:1. Used for text-to-video only.
generate_audio: Toggle native audio on or off. Dialogue in quotation marks, plus ambient sound and effects.
negative_prompt: Optional. Describe what to exclude from the generation.

What it’s good for

Marketing and advertising videos
Social media content with dialogue
Multi-scene narratives and short stories
Product demonstrations
Cinematic sequences with synchronized audio

Limitations

Maximum 15 seconds per generation
Audio works best in English and Chinese
Character appearance can vary across separate generations
Complex physics interactions may not look fully natural
For longer videos, generate multiple clips and stitch them together

FAQ

What can Kling 3.0 generate on Soup.video?: Kling 3.0 turns a text prompt or still images into an MP4 clip from 3 to 15 seconds. Skip images for text-to-video. Add a start frame (and optionally an end frame or character elements) for image-to-video. Multi-Scene builds up to 5 shots in one generation. Native audio is optional and includes lip-synced dialogue.
How do I use Kling 3.0?: Write a prompt in the Director bar. Leave images empty for text-to-video, or attach a start frame to animate a still. Choose Single or Multi-Scene, then set Standard / Pro / 4K, duration, aspect ratio, and audio. Cook. You can close the tab; Soup.video emails you when the clip is ready.
How does text-to-video work?: Do not attach a start frame, end frame, or character images. The prompt becomes the whole shot. Set ratio, mode, and duration, then cook.
How does image-to-video work?: Upload a starting image and describe the motion. You can also add an end image so the clip lands on a specific last frame. Aspect ratio follows the still.
How does multi-shot work?: Turn on Multi-Scene. Each scene has a prompt and a duration (minimum 1 second). You can define up to 5 scenes. The sum of shot lengths must equal the clip duration, between 3 and 15 seconds.
Does Kling 3.0 generate audio?: Yes. Turn audio on to generate dialogue with lip sync, sound effects, and ambient sound in the same pass. Audio works best in English and Chinese. On Pro, voice control is billed at a higher per-second rate.
What modes and prices does Kling 3.0 support?: Standard, Pro, and 4K, billed per second. Standard is ✦ 12/s without audio and ✦ 18/s with audio. Pro is ✦ 16/s, ✦ 25/s with audio, and ✦ 29/s with audio plus voice control. 4K is ✦ 60/s with or without audio. A 3-second Standard clip without audio is ✦ 36.