Text to video
Describe a shot in a prompt and get a 4–15 second clip.
The H3 workflow is deliberately focused: every option maps to a documented MiniMax H3 setting, with no unrelated template models or unverified benchmarks.
Open MiniMax H3 generator2modes
Text and image to video
4–15s
Any whole second
768P· 2K
Two resolutions
6ratios
From 21:9 to 9:16
Describe a shot in a prompt and get a 4–15 second clip.
Start from your own photo and set a first frame, a last frame, or both.
Text-to-video uses minimax-h3/text-to-video. Image-to-video uses minimax-h3/image-to-video.
Choose an integer duration from 4 through 15 seconds.
Choose 768P or 2K.
21:9, 16:9, 4:3, 1:1, 3:4, or 9:16.
Image-to-video frame URLs come only from assets registered to the signed-in account under the existing upload boundary.
The browser cannot choose the provider task owner, generation price, settlement state, or refund state.