ModelsVideo modelsKling 3.0 Omni AI Video Generator
Multimodal video creation

Kling 3.0 Omni AI Video Generator

Kling Video 3.0 Omni is the reference-oriented member of Kuaishou’s 3.0 video model series, bringing multimodal context into video creation and editing.

Kling 3.0 Omni
KLING 3.0 OMNI EXAMPLES

Kling 3.0 Omni Examples

Explore Beverage Promotion from Three References, FPV Motorcycle Chase, Basketball Performance Scene. Each result is paired with its actual English prompt, saved case, and available reference inputs.

Reference 1
Reference 2
Reference 3

Beverage Promotion from Three References

Use three supplied visual references for a beverage promotion. The brief asks for shot changes rather than camera movement. The three @{...} markers identify this saved case’s reference images; they are not reusable IDs for your own uploads. For a new request, upload your own references and state each image’s role instead of retaining these saved-case markers.

Use @{20260612_1_c81e555d-802a-4e36-a983-f2fec0d2e43d}@{20260612_1_346ff0cb-9155-4c27-a450-dc6c5abe41f7}@{20260612_1_e14479f3-bf6f-4448-beda-c150a9f22719} to generate a beverage promotional video. No camera movement is needed; use shot transitions instead.
Use This Prompt

FPV Motorcycle Chase

The prompt maps a camera journey over a neon city: dive toward a motorcycle, follow it around obstacles, swing into a side orbit, then pull up to a top-down view. Read the timed instructions alongside the result; they do not establish that every movement is reproduced exactly.

"[0s] A first-person FPV drone shot above a neon cyberpunk city at night, holographic ads, glowing signs, steam vents, and wet reflective streets below. [2s] The camera dives sharply into the urban canyon and locks onto a futuristic motorcycle speeding through the streets. [4s] It chases close behind the bike, weaving around elevated signs, flying under bridges, and banking hard around corners. [6s] The camera swings outward into a smooth side orbit as the motorcycle drifts through an intersection, sparks and mist filling the frame. [7s] It accelerates forward past the rider and pulls up into a dramatic top-down reveal of the glowing city grid. Immersive FPV pursuit, dynamic chase camera, neon reflections, fast banking turns, low-altitude urban flight, premium cyberpunk cinematic energy."
Use This Prompt

Basketball Performance Scene

Create a basketball scene on an indoor court, with the athlete’s performance supplying the visual action.

"A cinematic basketball scene on a professional indoor court. A tall muscular basketball player resembling an NBA superstar (yellow jersey, number 22 style, no logos) is confidently facing anthropomorphic dog(1 vs 1). The dog toned figure, pure white short fur, Blue jersey, focused, dribbling naturally with surprising skill. Crowd in the background, stadium lights bright, polished wooden court reflecting motion. [00:00-00:02] Close-up: the player smirks, relaxed stance. Cut to the dog gripping the ball, low angle, determined expression. [00:02-00:05] 1v1 begins. The dog dribbles fast, low to the ground. Quick crossovers, sudden direction changes. The player reacts but slightly off balance. [00:05-00:07] The dog performs an unexpected ankle-breaker move. The player stumbles slightly, crowd reacts. Dynamic camera shake, fast motion blur. [00:07-00:09] The dog drives past the player, jumps with exaggerated athletic motion. Slow motion: mid-air dunk or clean layup. [00:09-00:10] Ball goes in cleanly. The dog lands confidently. The player turns, surprised. Crowd erupts."
Use This Prompt
WHAT IS KLING 3.0 OMNI?

What Is Kling 3.0 Omni and What Is It Useful For?

Kling Video 3.0 Omni is the reference-oriented member of Kuaishou’s 3.0 video model series, bringing multimodal context into video creation and editing. Omni is relevant when subject references and directed scene transformation are central to the brief. The beverage campaign assigns three visual inputs, while the saved person-replacement request illustrates an editing task with a video input in its case details.

Multimodal Creative Direction

Omni belongs to Kling’s unified multimodal 3.0 model family. References can contribute a subject or other visual direction to the requested scene.

Audiovisual Scene Creation

The 3.0 release includes native audio-video output. The actual mode and generated clip determine whether the requested sound is present.

Subject-Led Sequences

The model family supports connected shots and reference-led subjects. The input roles matter when the same subject should carry through a sequence.

WORKFLOWS

Choose Your Starting Input

Describe the scene you want to create, then add image or video references when needed. Explain how each reference should guide the request.

Direct a Scene with References

Add the references needed for your request and explain their roles. For person replacement, the source video provides the scene and movement, while the person image provides appearance. Follow the displayed upload limits.

Start with a Text Brief

Describe the subject, setting, action, and camera movement in the prompt. Reference inputs are optional in the current form; use them when an existing subject or scene should guide the request.

PROMPT GUIDE

How to Direct Kling 3.0 Omni with a Prompt

Learn from the actual Kling 3.0 Omni requests below. Each card explains a specific instruction, shows its result and available inputs, and keeps the complete original English prompt. Adapt the subject and reference roles to your own task.

Reference 1

Assign the Source Video and Replacement Person

Compare the source video, person reference, and generated result: the video provides the scene and movement, while the image provides the replacement person’s appearance. In the original prompt, @{https://...mp4} identifies the source video and @{20260612_1_...} identifies the person image. For your own request, upload both files and explain their roles; copying the prompt does not attach or relink the files.

Replace the person in @{https://ai-video.weshop.ai/20260612_1_1dedd346-0b42-48aa-bbed-ffbb7eebbfda_810x1440.mp4} with the person in @{20260612_1_f695985f-2687-43a3-a193-8c2077d02c2a}.

Put the Events in Order

Describe what happens first and how the next action or reveal follows it. In this case: “Once complete, reality resumes dramatically faster than before.”

A person is about to perform a simple action when a giant floating loading bar suddenly appears above their head. Everything pauses while the loading slowly reaches 100%. Once complete, reality resumes dramatically faster than before. Realistic AR interface integrated into the environment, authentic reactions, funny timing, highly shareable TikTok concept.
Reference 1

Describe a Gentle Camera Push-In

Describe the gentle camera push-in requested by the prompt, then check the result’s framing. In this case: “Gentle camera push-in, soft film grain, warm highlights, natural shadows, subtle motion in hair, clothing, and light reflections, minimal but striking graphic movement, premium fashion-editorial vibe,”

Modern editorial YouTube intro animation with a bold lifestyle fashion aesthetic. Gentle camera push-in, soft film grain, warm highlights, natural shadows, subtle motion in hair, clothing, and light reflections, minimal but striking graphic movement, premium fashion-editorial vibe, cinematic, stylish, visually bold, smooth looping motion graphics, 4k resolution.
Reference 1
Reference 2

Give a Local Effect Its Target

The actual prompt is the short request “box add realistic explosion effects”. It identifies a target and an effect, but does not assign different jobs to multiple references or specify exact timing. Inspect the effect and the surrounding scene without inferring those missing controls.

box add realistic explosion effects
HOW TO USE

How to Use Kling 3.0 Omni on WeShop AI

Start with the Kling 3.0 Omni selection and the input path described below. Review the available settings and points before submitting.

  1. Provide the Input for This Model

    Enter the scene and action, then upload any image or video references needed for your request. For video editing, include the source clip and describe the change. Explain each reference’s role before generating.

  2. Check the Available Settings and Points

    Durations: 3–15 seconds. Aspect ratios: 16:9, 9:16, 1:1. These are selectable settings, not fixed pixel dimensions or file-format guarantees. Check the original output for the actual dimensions and duration. Review the displayed points and your plan’s model access before submitting.

  3. Open and Check the Result

    Compare the referenced person or product across shots, and review replaced regions and transitions. Inspect delivered sound separately; the Omni label is not a guarantee of exact identity or timing.

Kling 3.0 Omni Tutorial

Follow this Hotel Lobby tutorial for reference roles, the original prompt, case settings, and result checks.

MODEL SELECTION

Which Kling Selection Fits Your Task?

Kling 3.0 supports both text-to-video and image-to-video. Choose text for a new scene, or images when an existing composition should guide the shot. Consider Omni for reference-led creation and video edits, or O1 for its earlier multimodal workflow.

DimensionKling 3.0Kling 3.0 OmniKling O1
Model roleNative audiovisual generation and shot directionUnified multimodal direction and subject referencesEarlier unified multimodal generation and editing model
WeShop starting inputsText; opening image; first and last imagesText; source video and reference imagesReference: text with optional images or video; Image to Video: opening image and optional ending image
Current WeShop options3–15 seconds; Generate Audio on/offDuration: 3–15 seconds3–10 seconds in both modes; 9:16, 16:9 or 1:1 in Reference mode
Model role
Kling 3.0
Native audiovisual generation and shot direction
Kling 3.0 Omni
Unified multimodal direction and subject references
Kling O1
Earlier unified multimodal generation and editing model
WeShop starting inputs
Kling 3.0
Text; opening image; first and last images
Kling 3.0 Omni
Text; source video and reference images
Kling O1
Reference: text with optional images or video; Image to Video: opening image and optional ending image
Current WeShop options
Kling 3.0
3–15 seconds; Generate Audio on/off
Kling 3.0 Omni
Duration: 3–15 seconds
Kling O1
3–10 seconds in both modes; 9:16, 16:9 or 1:1 in Reference mode

Choose Kling 3.0 When

Create a scene from a written brief, animate an image, or guide both ends of a shot. Choose Kling 3.0 for 3–15-second requests with optional generated sound.

Choose Kling 3.0 When

Choose Kling 3.0 Omni When

Your task depends on reference subjects or a change to an existing video. Open Omni to prepare the source clip and references for that workflow.

Choose Kling 3.0 Omni When

Choose Kling O1 When

Choose O1 Reference for a written brief or image/video references, or Image to Video for opening and ending frames. Use O1 when its 3–10-second range fits your scene and workflow.

Choose Kling O1 When
MORE VIDEO MODELS

Explore More Video Models

Open a related model page to review its capabilities, examples, and current generation options.

FAQ

Kling 3.0 Omni FAQ

Answers about this version, its inputs, and how to get started.

What is Kling 3.0 Omni?

Kling Video 3.0 Omni is the reference-oriented member of Kuaishou’s 3.0 video model series, bringing multimodal context into video creation and editing.

What is Kling 3.0 Omni useful for?

Omni is relevant when subject references and directed scene transformation are central to the brief. The beverage campaign assigns three visual inputs, while the saved person-replacement request illustrates an editing task with a video input in its case details.

What inputs can I use? Can I upload audio?

Use a text description with optional image and video references. For person replacement, provide the source video and a person image. This entry does not support uploading audio files. Generate Audio controls whether the output video includes sound.

Which duration and format settings are available?

Durations: 3–15 seconds. Aspect ratios: 16:9, 9:16, 1:1. These are selectable settings, not fixed pixel dimensions or file-format guarantees. Check the original output for the actual dimensions and duration.

Is Kling 3.0 Omni free, and how is usage charged?

Check the current WeShop Pricing page for your plan’s model access and allowance, and review the points estimate in the generator for your selected settings and quantity. This page does not quote a fixed per-request price or transfer another provider’s subscription quota to WeShop.

What limitations should I check?

Official model capabilities and available WeShop controls are separate; no unlisted editor or reference limit is promised. Compare the referenced person or product across shots, and review replaced regions and transitions. Inspect delivered sound separately; the Omni label is not a guarantee of exact identity or timing.

What do the Kling 3.0 Omni examples demonstrate?

The saved requests show Beverage Promotion from Three References, FPV Motorcycle Chase, Basketball Performance Scene. Read each actual prompt and reference assignment alongside its result. These individual examples are not a success-rate measurement or a guarantee of exact preservation.

How does Kling 3.0 Omni differ from related models?

Compare Omni for reference-led creation and editing with standard Kling 3.0’s frame-led entry and the earlier O1 workflow. Their actual input paths matter more than assuming all 3.0-family controls are interchangeable.

How should I write a Kling 3.0 Omni prompt?

The actual request asks to replace a person in a specified video with a person from another reference. The video supplies the scene and motion; the person reference supplies appearance. Open the case for those inputs, and confirm the execution form accepts them before reuse. Read the complete paired prompt and actual inputs in the guide, then adapt them to your own task.