metadata
extra_gated_eu_disallowed: true
extra_gated_fields:
Country: country
I agree to the MiniMax H3 Community License Agreement and Acceptable Use Policy: checkbox
license: other
license_name: minimax-h3-community-license-agreement
license_link: https://huggingface.co/MiniMaxAI/MiniMax-H3/blob/main/LICENSE
base_model: MiniMaxAI/MiniMax-H3
tags:
- minimax-h3
- lora
- 3d-animation
- text-to-video
- image-to-video
widget:
- text: Girl with a telescope at dusk
output:
url: samples/20260926-210654-933_3670611240.mp4
- text: Raccoon and hedgehog in a garden
output:
url: samples/20260926-213610-328_502243029.mp4
- text: Otter floating with an apple
output:
url: samples/20260926-222120-863_3795468811.mp4
- text: A woman holding a cup in a cafe
output:
url: samples/20260927-175329-712_377046177.mp4
3D Animation Style LoRA for MiniMax H3
Trigger: 3dsrx
Prepared for T2V using the previously created 3d_animation dataset. The learned 3D animation style also carries over to I2V workflows.
Recommended Strength
Text-to-Video: 0.8
Image-to-Video: 0.4
For text-to-video generations featuring human subjects, a LoRA strength of 0.8 may help maintain prompt adherence and character consistency.
For animals, fantasy characters, and environments, try a strength of 1.0 for a more pronounced 3D animation style.
- Prompt
- Girl with a telescope at dusk
- Prompt
- Raccoon and hedgehog in a garden
- Prompt
- Otter floating with an apple
- Prompt
- A woman holding a cup in a cafe
License
This LoRA is distributed under the MiniMax H3 Community License Agreement. See LICENSE for the complete agreement and Acceptable Use Policy, and NOTICE for the attribution and modification notice.