Plan a shot with MiniMax H3
Use MiniMax H3 in Vidma.AI for text-to-video, image-to-video. Describe who or what is on screen, what changes over time, and how the camera should move. Where supported, use images or video to guide the result. Generated video can reinterpret motion, faces, lettering and continuity, so inspect the clip before using it.
- 2 available workflows
- Model-specific input limits
- Settings shown before submission
Choose the right starting point
Build a shot from words
Start with a scene description. Name the subject, action, camera movement and visual style. Selectable resolutions: 2k. Select an available aspect ratio in the workspace. Choose from the duration options shown in the workspace.
Animate a still image
Add 1 reference image. Describe the movement and shot you want to build from the still. Selectable resolutions: 2k. An aspect-ratio choice may not be available for these inputs. Choose from the duration options shown in the workspace.
How to use MiniMax H3 in Video Studio
01 / Open the model
Use the creation button to enter Video Studio with this model and its default workflow selected. Sign in if prompted.
02 / Set the scene
Choose one of the supported workflows. Add only the image or video references requested by that mode. Write a concise prompt that describes the subject, action, composition and movement. Use media you have permission to upload.
03 / Review the controls
Select only the resolution, aspect ratio and duration options the workspace offers for this model and mode. Check the estimated credit cost before submitting.
04 / Inspect the clip
Review movement, continuity, faces and any on-screen text. Save a useful result or adjust one part of your brief and try again; exact reproduction is not guaranteed.
MiniMax H3 FAQs
What can I use with MiniMax H3?
The current Vidma.AI DEV catalog lists text-to-video, image-to-video. Available choices can change; the Video Studio controls are the source of truth when you create.
Do I need an image or a video?
Not for text-to-video. Other supported workflows may require reference images or source clips; see the per-workflow limits above.
Which output settings can I change?
Resolution, aspect ratio and duration controls depend on the selected workflow and inputs. The per-workflow details above list the controls declared by the current backend.
How much does a generation cost?
The estimate appears in Video Studio before submission. It can vary with model, workflow and settings; this page does not promise a fixed or free price.
Will the clip exactly match my brief?
Not necessarily. Review motion, continuity and fine details before using an output. Revise the prompt or inputs if the first result misses an important part of your idea.