Upload an image and describe the motion, camera work, and atmosphere you want. Wan 3.0 creates high-quality AI videos up to 30 seconds and 1080p.
Wan 3.0 requires an uploaded image to generate a video. Text-only generation is not currently supported.
Wan 3.0 is the next-generation model in the Wan AI video series, designed for longer visual narratives, stronger consistency, and richer audiovisual results. It interprets the relationship between a reference image and prompt so creators can direct subject motion, camera movement, scene changes, and pacing more precisely.
LitVideo currently offers Wan 3.0 for image-to-video generation. Start with an image to define the subject and composition, or add an end frame to guide the final shot. Generate videos up to 30 seconds and 1080p.
Upload a start image and optionally add an end frame.
Choose a short motion clip or a complete longer shot.
Choose 480p, 720p, or 1080p output.
Add sound that matches the atmosphere of the scene.
A strong reference image establishes the visual foundation for a person, product, or scene. Wan 3.0 then extends it through time, motion, camera direction, and sound based on your prompt.
Set anything from a 2-second motion test to a complete 30-second shot. Longer durations give product showcases, staged actions, and gradual environmental changes more room to unfold without stitching together multiple generations.
Choose 480p, 720p, or 1080p based on whether you are testing an idea, publishing to social media, or preparing footage for editing. Preview motion at a lower quality, then render the final version in 1080p.
With only a start image, Wan 3.0 extends the scene according to your prompt. Add an end image to specify the final composition, subject position, or product state for a more directed transition.
Enable audio to generate sound that suits the scene. Describe rain, footsteps, ambient sound, or other audible details in your prompt to give the finished video a stronger sense of presence.
Start with a clear product image and use a prompt to direct the camera, lighting, and environment. Wan 3.0 turns those instructions into a smooth product video.
Reference image
Prompt
Generated video
Choose a clear, well-composed image with a visible subject. To control the ending, add an end frame in the full editor.
Describe how the subject moves, how the camera travels, how the lighting changes, and which people or objects must stay consistent.
Choose the video length and 480p, 720p, or 1080p output, then generate and preview the result.
You do not need to repeat details already visible in the image. Focus on what happens next and how the camera should capture that change.
Recommended prompt formula
Subject motion + camera movement + environmental change + lighting and mood + consistency requirements + duration
Portrait close-up
The person looks into the camera, smiles gently, and then shifts their gaze to the right. The camera pushes in very slowly from a medium close-up to a facial close-up. Keep the background softly blurred as the natural window light becomes slightly brighter and the hair moves subtly in the breeze. Preserve the face, hairstyle, and clothing. Natural motion, cinematic, 10 seconds.
Best for portraits, character concepts, and personal story shots. Restrained movement helps keep facial details stable.
Product showcase
Start with a front view of the product. The camera moves smoothly in a 90-degree orbit to the right while slowly pushing closer. Fine highlights travel across the metal edges, reflections shift naturally across the tabletop, and the background lighting gradually changes from cool to warm. Keep the logo, shape, proportions, and surface materials completely consistent. Premium commercial look, 8 seconds.
Best for ecommerce hero visuals, product ads, and branded social content. State which product details must not change.
Landscape scene
The camera glides slowly forward just above the lake. Mist gradually clears from the distant valley, subtle ripples form on the water, and leaves move gently in the wind. Morning sunlight slowly emerges from behind the clouds and illuminates the mountain ridge, shifting the overall palette from cool to warm. Preserve the terrain and major landmarks. Calm, cinematic, 15 seconds.
Best for concept environments, travel content, and mood films. Limit a longer shot to one or two major changes for steadier pacing.
Start-to-end transition
Begin with the pencil architectural sketch in the start frame. The lines gain detail and gradually transform into realistic building materials. Keep the camera at the same angle while pushing forward slowly as the walls, glass, and surrounding landscape take shape in sequence, arriving naturally at the complete building shown in the end frame. Keep the perspective, composition, and architectural outline consistent. Smooth continuous transformation, 12 seconds.
Best for clear before-and-after transitions such as a design becoming a finished product, day turning to night, or packaging changing state.
The start frame defines where the video begins, while the end frame gives the generation a clear destination. Whether you need one image or two depends on whether you want to control the direction of motion or the complete visual transition.
Let Wan 3.0 extend motion, camera work, and environmental changes from one image and a prompt. Describe what happens after the image rather than forcing a fixed endpoint.
Best for: Subtle portrait motion, product orbits, animated scenery, and shots that benefit from more generative freedom.
Define the opening and closing shots with two compositionally compatible images, then explain how the transition happens in the prompt. Keep the aspect ratio consistent and avoid extreme changes in subject position.
Best for: Before-and-after changes, directed camera endpoints, character repositioning, and stories with a clearly defined result.
From product showcases to character performances, a clear start image makes the direction of generation easier to control.
Add orbiting camera moves, material reflections, and natural environmental motion to a still product image while preserving its appearance and proportions.
Add subtle expressions, body movement, and continuous camera work to portraits, character designs, or storyboards.
Extend the weather, lighting, and spatial movement in a concept image to build a richer mood and narrative rhythm.
Turn a key visual into engaging Reels, Shorts, product ads, or event teasers.
Both models can create high-quality AI video. In the workflows currently available on LitVideo, Wan 3.0 is better suited to longer image-to-video shots and flexible output quality. The specifications below reflect the current LitVideo model settings.
Comparison
Wan 3.0
Wan 2.7
LitVideo input
Image to video with start and end frames
Text to video, image to video, and reference modes
Video length
2-30 seconds
2-15 seconds
Output quality
480p, 720p, 1080p
720p, 1080p
Best suited to
Longer narratives, complete product showcases, and start-to-end transitions
Shorter shots, text-first ideation, and varied input workflows
Check your reference image, intended duration, and output use before generating. A clear plan reduces unnecessary retries and helps the result match your goal.
Choose an image with a complete subject, clear details, and minimal obstruction. The more visible the face, product logo, or key object is, the more your prompt can focus on motion and camera direction.
Use a shorter duration for one action or a simple camera move. Increase it to 10-30 seconds for staged changes, slower pacing, or start-to-end transitions. Avoid packing too many events into a short clip.
Wan 3.0 credit usage depends on the selected output quality and video length. The generator displays the current credit requirement as you adjust the settings.
Wan 3.0 is the next generation of the Wan AI video model series. LitVideo currently offers an image-to-video workflow that turns a reference image into a dynamic video up to 30 seconds long.
Text-only generation is not currently supported. Upload an image first, then use a prompt to describe the motion, camera work, lighting, and atmosphere.
Wan 3.0 on LitVideo supports videos from 2 to 30 seconds. Choose the length based on the complexity of the shot, narrative pacing, and intended use.
You can currently choose 480p, 720p, or 1080p. Use 1080p for publishing or further editing.
Yes. The start image establishes the subject and composition, while an end image can define the final state the shot should reach.
Focus on what should change in the image: subject motion, camera movement, environmental response, lighting, and pacing. Also state which people, products, or background details must remain consistent.
Generating a Wan 3.0 video on LitVideo uses credits. The required amount depends on the video length and the selected 480p, 720p, or 1080p quality. The generator shows the current credit requirement before you begin.
Use a clear, stable composition with the full subject visible. Avoid hiding important facial or product details. When using start and end frames, keep the same aspect ratio and a consistent subject appearance.
In the current LitVideo settings, Wan 3.0 image to video supports up to 30 seconds with 480p, 720p, and 1080p output. Wan 2.7 supports up to 15 seconds at 720p or 1080p and also offers text-to-video and reference workflows.
Yes. When audio is enabled, Wan 3.0 can generate sound that matches the scene. You can also describe ambient sound, footsteps, rain, or other audible details in the prompt.
Upload an image, direct the motion and camera, choose a duration, and create your Wan 3.0 image-to-video clip.
