What remains fixed
- Product and campaign stills
- Illustration animation
- Character and fashion shots
- Controlled first-frame composition
MiniMax H3 Image to Video Generator
Animate a source image into a directed 2K video. Use the opening frame as the visual anchor, add an ending frame when you need a controlled transition, and let the prompt define movement, camera behavior, atmosphere, and sound.
ReferenceFramesFirst & Last FrameTextText to Video
ImagesAdd reference images0/9up to 9JPG/PNG/WebP · up to 9
VideosAdd reference videos0/3≤3 · 15sMP4/MOV · ≤3 · 15s total
AudioAdd reference audio0/3≤3MP3/WAV · ≤3 · needs image/video
Prompt
Upload up to 9 images, 3 videos, and 3 audio clips (12 references total). Use @Image, @Video, or @Audio in your prompt to identify each reference.
Inspiration
MiniMax H316:95s2K
MiniMax H316:95s2KGenerate MiniMax H3 Image-to-Video · Free
Keep the composition
When MiniMax H3 image to video gives you more control
Image-to-video begins with a frame you already approve. That makes it useful for product images, illustrations, campaign key art, architectural views, and character shots where text alone would introduce too much visual variation. On this page the frame workflow is preselected: upload the starting image, add a closing frame when the end state matters, describe only the motion and changes you want, and protect fixed details in the prompt.
Protect what must not change
Name the logo orientation, face, product silhouette, or background geometry that should remain stable during motion.
Leave movement space
A tight crop limits camera travel. Choose a source frame with useful negative space in the direction the subject or camera should move.
Use a final-state instruction
Describe where the subject and camera should settle. A clear ending often makes the whole movement feel more intentional.
A four-step workflow
From source image to intentional MiniMax H3 motion
Upload a clean starting image
Use a sharp image with an intentional subject, readable edges, and enough room for the movement you want.
Describe motion, not the whole image
Explain what moves, how the camera moves, what remains stable, and how the shot should end.
Add an ending frame if the final state matters
Use a closing frame for product reveals, before-and-after changes, or transitions where the final composition should be controlled.
Match the aspect ratio and inspect stability
Choose an output ratio that fits the source composition, then check faces, product geometry, text, and background edges before adding more style.
Motion prompt anatomy
Write the movement around the image
Prompt formula
Subject motion + camera motion + stable details + atmosphere + ending state + sound
Example prompt
The bottle remains centered and label-forward while condensation slowly forms. A controlled half-orbit camera move, subtle light sweep across the glass, dark studio background stays stable, ending on a clean hero frame with soft room tone.
Credits for frame-led creation
MiniMax H3 image-to-video pricing for 2K generations
Compare one-time credit packs for MiniMax H3 image-to-video and first-and-last-frame generation. Choose by credit volume and expected production frequency.
Clear answers
MiniMax H3 image-to-video questions, answered
How does MiniMax H3 image to video work?
You upload a source frame and write a motion prompt. The image anchors composition and appearance while the prompt directs movement, camera, atmosphere, and sound.
What image works best for image-to-video?
Use a sharp, well-composed image with clear subject separation, readable edges, and enough space for the intended movement.
Do I need a first and last frame for MiniMax H3 image to video?
Use a starting image when you want to animate a known composition. Add a last frame when the ending state must be controlled, such as a product reveal, pose change, or before-and-after transition.
Can I use a product photo or illustration?
Yes. Product stills, illustrations, campaign key art, architecture, and character frames are practical starting points when their composition should be preserved.
Is image-to-video the same as reference-to-video?
No. Image-to-video treats an image as the opening visual frame. Reference-to-video uses a subject image to guide recognizable identity while the scene can change more freely.