MiniMax H3 image to video

Use one first frame to direct a MiniMax H3 video.

When the opening composition matters, attach one approved first frame and describe only the movement that should develop from it. MiniMax H3 derives framing from that image and keeps 768P or 2K plus a 4 to 15 second duration range.

Animate with MiniMax H3Browse video workflows
A local first-frame motion reference used to discuss image-led direction; it is not presented as a MiniMax H3 result.
Start with
One approved first-frame image and a prompt
Framing
Derived from the uploaded image
Output
768P or 2K for 4 to 15 seconds

MiniMax H3 first-frame motion directions

These local references show how a stable first image can establish the subject and composition before the prompt directs movement. They are not presented as MiniMax H3 outputs.

Every card identifies whether it is a verified model-family example or a visual reference. Reference cards are not model benchmarks or before-and-after claims.

Stable first frame

Choose an image that already makes the subject, room, and viewpoint easy to identify.

Reference direction, not a model output

Local ImageStyle visual reference. It is not presented as a model-specific result.

Subject motion

Let the prompt describe the next action and pace instead of restating the entire source image.

Reference direction, not a model output

Local ImageStyle visual reference. It is not presented as a model-specific result.

Camera restraint

Use one explicit camera behavior to protect the image-led composition while motion develops.

Reference direction, not a model output

Local ImageStyle visual reference. It is not presented as a model-specific result.

What a MiniMax H3 first frame can establish

The opening composition

Use one approved frame when the subject placement, setting, or crop must be visible before motion begins.

A focused movement brief

Use the written prompt for the motion, camera behavior, and pacing that the image cannot define on its own.

A current model-aware delivery

The uploaded image determines framing, while the workspace offers the current 768P or 2K and 4 to 15 second controls.

How to make a MiniMax H3 video from an image

Treat the first frame as the visual anchor, then make the prompt about the change over time rather than recreating the image.

  1. 1

    Choose one authorized first frame

    Use a clear image that already establishes the subject, composition, and visible details that must remain recognizable.

  2. 2

    Describe motion instead of the still image

    Write what moves, how the camera responds, and how the scene should arrive at its ending state.

  3. 3

    Select output quality and duration

    Choose 768P or 2K and a 4 to 15 second duration. MiniMax H3 derives the frame from the upload.

  4. 4

    Inspect continuity before use

    Review the subject, motion, frame edges, visible text, and ending state before publishing the clip.

What to check before using a first-frame video

The image does not replace the prompt
MiniMax H3 image-to-video requires both one first frame and a written direction. Use the prompt to state the motion that must happen next.
Framing comes from the uploaded image
The image-to-video workflow does not expose a separate frame selector because the source image establishes the output framing.
Only use images you are allowed to animate
Confirm that you have permission to upload and publish the source image, then review the generated result before public use.

MiniMax H3 image-to-video FAQ

What is MiniMax H3 image to video?
MiniMax H3 image to video creates a short clip from one first-frame image plus a written motion direction. The uploaded image determines the framing, while the current workspace offers 768P or 2K and 4 to 15 seconds.
Does MiniMax H3 image-to-video require a prompt?
Yes. Attach one approved first frame, then write the movement, camera behavior, or scene change that should develop from it.
Can I select the aspect ratio after uploading an image?
No. The current MiniMax H3 image-to-video workflow derives framing from the uploaded image rather than exposing a separate aspect-ratio control.
How many images can I add to MiniMax H3?
The current MiniMax H3 image-to-video workflow accepts one first-frame image. Use it to establish the subject and composition before you direct motion.

Updated:

Capability source: Current ImageStyle MiniMax H3 video workspace.

Animate with MiniMax H3