Animate a still image with AI
AI Image to Video Generator
Upload portraits, product photos, illustrations, motion clips, or audio cues and describe how they should guide the result. Use @ mentions to connect a prompt instruction to a specific reference.
Choose a supported video model, output ratio, quality, and duration before generating. Seedance 2.0 accepts multiple references instead of treating the first two images as fixed first and last frames.
Animate images with supported AI video models
The Image to Video model list is filtered to reference-guided workflows. Quality, ratio, duration, and credit cost update with the selected model.
Seedance 2.0480p / 720p / 1080pSeedance 2.0 Mini480p / 720p
Veo 3.1720p
HappyHorse 1.1720p
Runway720p
Kling 2.6720p · 5/10 sec · Optional audio
Wan 2.7720p
Hailuo 2.3720p
Grok480p / 720pHow to turn an image into a video
- 01
Upload reference material
Add the images, videos, or audio that should guide identity, composition, motion, rhythm, or sound.
- 02
Describe the motion
Separate subject movement, camera movement, environmental motion, speed, and the intended ending state.
- 03
Choose settings and generate
Select a compatible model, ratio, quality, and duration, then review how closely the clip preserves the source.
Image to video motion prompt ideas
The source image already defines the look. Use the prompt to direct motion rather than redescribing every visible detail.
Prompt template
Portrait
Portrait with restrained movement
Small facial and camera movements usually preserve identity and composition better than several large actions at once.
Prompt direction
She looks toward the camera and blinks naturally, a light breeze moves a few strands of hair, slow camera push-in, calm expression, keep the face and background consistent.
Prompt template
Product
Product orbit
Keep the object fixed while directing a controlled camera path and a simple lighting change.
Prompt direction
The product remains centered and unchanged, slow clockwise camera orbit, a soft highlight travels across the surface, subtle background haze, end on the original front angle.
Illustration
Illustration brought to life
Animate hair, fabric, particles, and the camera while keeping the character design readable.
Prompt direction
Her hair and dress move gently in the wind as glowing butterflies pass around her, slow camera arc from behind to three-quarter view, preserve the character design and pastel sky.
Environment
Landscape atmosphere
Layer subtle environmental movement with one slow camera direction for a stable scenic clip.
Prompt direction
Clouds drift slowly from left to right, grass ripples in waves, chimney smoke rises and bends with the wind, gentle aerial push toward the house, preserve the vertical composition.
Which images can you animate?
Portraits and character art
Add restrained facial, hair, clothing, and camera movement to a clear single-subject image.
Product photography
Create short product reveals with controlled camera motion, lighting, and atmosphere.
Illustrations and concept frames
Introduce environmental motion and depth while preserving the original visual direction.
Storyboard frames
Test motion, pacing, and camera ideas from an approved composition before production.
How to write an image to video motion prompt
Tell the model what should move, how the camera should move, and what must remain stable. Motion language should complement the source frame.
Subject motion + camera motion + environment motion + speed + ending state
Separate three kinds of motion
Describe the subject, camera, and environment independently so the model can interpret each direction.
Match movement to the pose
A seated portrait, close product shot, and wide landscape each support different motion ranges.
State what should stay consistent
Name the face, product shape, composition, or background elements that should not change.
Image to video input and output settings
Seedance reference limits are validated before upload, while output ratio, quality, duration, and credits depend on the selected model.
- Reference images
- Up to 9 JPG, PNG, or WebP images; aspect ratio up to 2.5.
- Video and audio
- Up to 3 MP4/MOV videos and 3 MP3/WAV audio files; 2–15s each and 15s total per type.
- Prompt length
- Up to 5,000 characters for the motion prompt.
- Output settings
- Aspect ratio, quality, and duration update by model.
Learn more
Explore practical prompt guides, then return to the generator with a clearer brief.
Image to Video FAQ
What is image to video generation?
Image to video generation uses a still image as the visual starting point for an AI-generated clip. A motion prompt directs the subject, camera, and environment over time.
What images work best for image to video?
Use a clear image with a readable subject, intentional composition, and enough detail. Motion that fits the pose and framing is usually easier to preserve.
Will the generated video preserve my original image?
The model uses the image as its starting reference, but preservation varies with the model and requested motion. Smaller, specific movements generally reduce unwanted changes.
How should I describe camera and subject motion?
Describe them separately. For example, specify a slow subject turn, a gentle camera push-in, and light background movement, then state what should remain consistent.
Which reference formats are supported?
Seedance 2.0 accepts JPG, PNG, and WebP images, MP4 and MOV videos, and MP3 and WAV audio in the reference uploader.
Which aspect ratios and durations can I choose?
The available aspect ratios and durations depend on the selected video model. The generator updates the controls to show supported combinations.