Your generations and videos, together. Open a card to preview or download.
Generation guide
/ai-video-generator
Create with MiniMax H3 in one focused video workspace
MiniMax H3 Max is the quality-led H3 tier for stronger prompt adherence, visual finish, and intentional camera motion. Start from text or upload a first frame with an optional end frame, then choose 480P or 768P and any duration from 5 to 15 seconds.
H3 Max
480P / 768P
5–15 seconds
First & last frame
01
Write a shot, not a static image description
A MiniMax H3 prompt can describe subject movement, camera path, timing, changes in light, and the intended final composition together. Treating the prompt as a concise director brief reduces contradictory motion and gives the generated clip a more complete visual rhythm.
02
Use first and last frames as motion boundaries
Image-to-video requires a first frame and accepts an optional final frame. The first image anchors subject, product, environment, and canvas. The last image can define where the action should resolve. Output aspect ratio follows the first frame without a separate crop setting.
03
Separate exploration from delivery with two tiers
H3 Max Turbo is designed for rapid tests of prompt and motion direction. H3 Max is the quality-led choice for a more polished final pass. Both tiers share the same workspace and private history, so changing speed does not mean learning another control system.