Preview the upcoming MiniMax H3 audio-video model. The current form executes through Grok Imagine Video 1.5 in text-image-to-video mode by default. The default model has no audio capabilities.
Generate video · 6 credits total
Preview the MiniMax H3 audio-video generation workflow. The current form runs Grok Imagine Video 1.5 in text-image-to-video mode by default. Write a prompt that describes one visible subject, one action, and one camera move. The default model does not generate synchronized sound, dialogue, or music. Review the visual output for motion quality.
Describe one visible subject, one action, and one camera move. For example: 'A person walks down a rainy city street at night, holding an umbrella. Camera follows the person from behind.'
The model creates a short video based on your prompt. No audio is produced. The clip shows the described motion and camera movement.
Check that the subject moves as described, the camera follows the specified path, and the environment remains consistent throughout the clip.
A person walks down a rainy city street at night, holding an umbrella. Camera follows the person from behind. Cinematic, moody atmosphere.
The rain, umbrella, and person walking should be visible. The camera should follow from behind.
Specify 'the umbrella is held above the person's head and remains fully visible throughout'.
Add 'heavy rain with visible streaks and puddles on the ground' to your prompt.
Yes, replace 'rainy city street' with 'rainy park path with trees and a pond'.
