MiniMax H3 Max generates video from a text prompt or animates a still image supplied as the start frame. Adding a start frame selects image-to-video mode automatically. The image then determines the output aspect ratio, overriding the Aspect Ratio setting.
Video Resolution offers 480P, 768P and 1080P. 480P costs the least, 768P is the default, and 1080P is refined from 768P. Video Duration ranges from 5 to 15 seconds. Cost increases in proportion to duration.
Prompt Expansion controls how MiniMax rewrites your prompt before video generation. Balanced processes the prompt quickly. Quality takes around 30 extra seconds to expand it, which can help when the original prompt is short or lacks detail.
Add both a Start Frame and an End Frame to generate a video that transitions between the two images.
Example: a museum forecourt
MiniMax H3 Max text-to-video result: a five-second camera move across a landscaped museum forecourt.
Pick a model and start generating
These models are tuned for detail — open one and put this guide to work.
