Create short scenes from a written brief or visual references with Kling O1, Kuaishou’s unified video generation and editing model. Choose Reference for text, images or video guidance, or Image to Video for opening and ending frames.
See a football-pub celebration, an outfit transition and a product demonstration. Play each clip, compare its reference images and copy the complete prompt to adapt it to your own scene.
Football Pub Celebration
A portrait supplies the woman at the centre of a fictional football-pub celebration. The prompt adds cheering fans, confetti and camera movement. Compare the reference with the woman’s appearance as the scene develops.
Inside a legendary football pub, a woman proudly holds a football and a glass of beer. The camera slowly pushes through rows of vintage football memorabilia. Suddenly every football on the shelves begins spinning. Historic World Cup moments appear as holographic projections across the bar. Fans from different countries gather behind her, raising glasses and cheering. The atmosphere intensifies with chants, drums and celebrations. Golden confetti rains from above as the camera rotates around her. Rich cinematic lighting, nostalgic football culture, World Cup night celebration, emotional storytelling, ultra realistic.
The first image supplies a clothing-and-accessories layout; the second supplies a person wearing the intended look. The prompt connects them through camera movement. Watch the transition from separate items to the modelled outfit.
The transition from the clothes in Figure 1 to the outfit in Figure 2 was achieved through camera movement.
An uploaded cosmetic organiser anchors a short product demonstration. The clip shows hands using its cover and drawer. Compare the compartments and product shape with the reference before using a similar result in a product listing.
Create a premium Shopify product usage tutorial video using the uploaded product as the hero subject. Preserve the product exactly as provided, including its shape, proportions, materials, colors, branding, packaging, and every original design detail.
Begin with a clean hero shot introducing the product, followed by a clear step-by-step demonstration of how to use it correctly. Each action should be natural, easy to understand, and filmed from the best viewing angle. Alternate between medium shots showing the complete process and macro close-ups highlighting important details, interactions, and product features.
Use bright natural lighting, realistic environments, smooth camera movements, and premium commercial cinematography. The pace should feel calm, intuitive, and instructional while maintaining a high-end Shopify brand aesthetic.
No text overlays, subtitles, promotional graphics, watermarks, or exaggerated visual effects.
Ultra realistic, premium Shopify tutorial video, commercial quality, 4K.
Kling O1 is Kuaishou’s unified multimodal video model for generation and editing. It combines a written brief with visual context, making it useful for reference-led product scenes, outfit transformations and short visual stories. In WeShop, choose between Reference and Image to Video according to the inputs you have.
Direct Scenes with Visual References
Use a person, product or scene image to guide the subject of a new clip. The football-pub and organiser examples show how an uploaded image can anchor a different scene or action.
Describe a Visual Transformation
Kuaishou describes O1 editing tasks such as changing clothing, modifying a background and restyling a scene with written instructions. State what should change and what should remain; inspect the result for unintended changes.
Connect a Starting and Ending Look
O1 includes first-and-last-frame generation. The outfit example pairs a clothing layout with a person wearing the target look, giving the transition a defined visual starting point and destination.
WORKFLOWS
Choose Reference or Image-to-Video Mode
The Model menu has two choices. Use Kling Video O1 Reference for a written scene with optional image or video references; use Kling Video O1 Image to Video for an opening image, with an ending image for a two-frame transition.
Reference Mode: Start with a Text Brief
Select Kling Video O1 Reference and describe the subject, setting, action, and camera direction. Add reference media only when it should guide the request.
Reference Mode: Add Image or Video Guidance
Keep Kling Video O1 Reference selected, add the media your request needs, and explain what each reference supplies, such as the subject, scene, or movement.
Image-to-Video Mode: Animate a Starting Frame
Select Kling Video O1 Image to Video, upload the opening image, and describe the action or camera movement that should follow it.
Image-to-Video Mode: Connect Two Frames
In Kling Video O1 Image to Video, supply the opening and ending images and describe the transition between them. Review the settings shown for this mode.
PROMPT GUIDE
How to Direct Kling O1 with a Prompt
Use the paired prompts and videos to study how a scene is specified: reference roles, an action or transformation, a setting and a visual style. Copy the full prompt, then adapt the subject and action to your own inputs.
Use an Interaction to Trigger the Outfit Change
This prompt links floating clothes, a cursor selection and an outfit change in a studio scene. Describe the trigger and the transformation together. When reviewing a similar clip, check the person’s face and the changing fabric.
The model stands in a clean studio. Around her, multiple floating clothing items hover in space like a fashion orbit. Each clothing piece gently floats and rotates. A cursor clicks one outfit. Immediately, the selected clothing flies toward the model and seamlessly becomes her outfit. Fabric naturally adapts to body shape, lighting remains consistent, identity unchanged. The surrounding clothes continue floating softly in the background.A single consistent fashion model is used throughout the entire video (identity locked). The video is a premium AI fashion transformation world where interactive elements float around the model. Each feature is triggered by subtle clicks. No software UI, no dashboard, only elegant floating interactive objects. Luxury editorial photography style, white seamless studio environment, soft lighting, high-end ecommerce campaign aesthetic. Every interaction instantly transforms reality.
Give the Scene Participants Specific Activities
The seagull-market prompt combines a harbour at dawn with birds inspecting, trading and displaying fish. These verbs define the intended activity; the morning light and documentary style define the visual treatment.
Long before the first fishermen arrive, seagulls secretly operate their own bustling seafood market on a quiet harbor. Fresh fish are inspected, traded, carefully displayed on tiny wooden stalls, and auctioned under the first golden light of dawn. Ocean mist, cinematic harbor atmosphere, realistic bird behavior, observational documentary style.
Build a World with Setting and Props
The squirrel-bank prompt defines an oak-tree interior, acorns as currency, vaults and miniature clerks. Use the setting, props and participants together to explain the fictional scene you want.
Hidden inside the hollow of an ancient oak tree is an elegant banking institution run entirely by squirrels. Acorns are counted like gold coins, tiny vaults protect winter savings, miniature clerks stamp paperwork, and woodland customers patiently queue for seasonal loans. Hyper-detailed forest interiors, warm autumn lighting, whimsical realism, cinematic storytelling.
Anchor a Short Story in One Visible Event
This story begins with a phone notification and a woman’s reaction, then asks for a longer chain of events. For your own short clip, make the main visible event clear and review which actions fit the selected duration.
A cinematic AI short drama. A young woman receives an unexpected phone notification: "Your ex has updated your personality." At first she laughs it off, but throughout the day she unconsciously starts speaking differently, ordering coffee she never liked, choosing clothes she would never wear, and repeating phrases her ex always used. She slowly realizes she's losing herself. Panicked, she opens the settings and desperately presses "Undo." The interface calmly responds: "Premium Required." The screen cuts to black. Natural acting, handheld iPhone cinematography, subtle UI overlays, realistic performances, modern fashion styling, emotionally unsettling, Black Mirror meets TikTok storytelling.
HOW TO USE
Create Your First Kling O1 Video
Choose a mode, describe the clip and review the settings before generating on this page. Open the returned result to inspect it before downloading or sharing.
Provide the Input for This Model
Choose Reference or Image to Video in the Model menu. Assign clear roles to image or video references in Reference mode; provide an opening image and an ending frame when needed in Image to Video mode. Check the settings again after switching.
Check the Available Settings and Points
Both modes offer 3–10-second requests. Reference mode also offers 9:16, 16:9 and 1:1 aspect ratios; Image to Video uses the supplied frame inputs. Review the points estimate for the selected mode, duration and quantity, and check your account’s model access before submitting.
Open and Check the Result
Generate the clip, then open the returned result. Check the person or product throughout the video, the transition and any visible text. Download the original file to check its actual dimensions and duration before delivery.
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.
Dimension
Kling 3.0
Kling 3.0 Omni
Kling O1
Model role
Native audiovisual generation and shot direction
Unified multimodal direction and subject references
Earlier unified multimodal generation and editing model
WeShop starting inputs
Text; opening image; first and last images
Text; source video and reference images
Reference: text with optional images or video; Image to Video: opening image and optional ending image
Current WeShop options
3–15 seconds; Generate Audio on/off
Duration: 3–15 seconds
3–10 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 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.
Answers about this version, its inputs, and how to get started.
What is Kling O1?
Kling O1 is Kuaishou’s unified multimodal video generation and editing model. This page is for the video model; use Reference for a written brief with optional visual references, or Image to Video for frame-led generation.
What is Kling O1 useful for?
Use it for short scenes built around a person, product or setting, outfit transitions and visual transformations. The examples include a fictional pub celebration, a clothing transition and a cosmetic-organiser demonstration.
What inputs do I need?
For text-only creation, select Kling Video O1 Reference and write a prompt. In that mode you can also add image or video references and explain their role. For an opening image or a first-and-last-frame transition, select Kling Video O1 Image to Video and supply the corresponding frame images.
Which duration, aspect ratio and resolution settings are available?
Both O1 modes offer 3–10-second requests. Reference mode provides 9:16, 16:9 and 1:1 aspect ratios. The Image to Video form provides opening and ending image inputs instead of an aspect-ratio menu. This form has no separate resolution selector; check the downloaded original for its pixel dimensions.
How should I write a Kling O1 prompt?
Identify the person, product or setting and explain what each reference supplies. Describe the action or transformation, then the camera and visual style. Read the full paired prompts in the guide: the fashion example uses a cursor selection to trigger the requested outfit change.
How do I choose between O1, Kling 3.0 and Kling 3.0 Omni?
Choose O1 for its text, reference or frame-led workflows when 3–10 seconds fits your clip. Consider Kling 3.0 for its opening-image and first-and-last-frame entry up to 15 seconds, or Kling 3.0 Omni for text and reference-led direction up to 15 seconds. Review the input controls on the page you choose.
Is Kling O1 free, and how is usage charged?
Review the current WeShop Pricing page for your plan’s model access and allowance. Before generating, check the points estimate for the selected mode, duration and quantity. Changing the request settings can change the estimate.
What should I check in the result?
Check faces, clothing and product details throughout the clip, especially during transitions or replacement. Review visible text and whether the requested actions fit the clip. For publication, use the original downloaded file and inspect its image and sound.
Can I choose a voice or request 4K output?
The current O1 form has no voice, audio-generation or 4K resolution selector. Writing dialogue or “4K” in a prompt does not add one. Check the original file’s properties and sound before using it.
How can I use the examples?
Play a clip and compare its reference images with the result. Copy Prompt copies the complete prompt; Use This Prompt opens the linked case. Adapt the prompt and replace reference images with your own materials.
