Image-Text-to-Video
MiniMax H3
Diffusers
Safetensors
text-to-video
image-to-video
video-to-video
text-to-audio-video
image-to-audio-video
image-text-to-audio-video
video-to-audio-video
audio-to-audio-video
audio-video-generation
multimodal
synchronized-audio-video
reference-to-audio-video
Instructions to use MiniMaxAI/MiniMax-H3 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use MiniMaxAI/MiniMax-H3 with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("MiniMaxAI/MiniMax-H3", dtype=torch.bfloat16, device_map="cuda") prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k" image = pipe(prompt).images[0] - Inference
- Notebooks
- Google Colab
- Kaggle
H3 and H3 Flex on Mac
#121
by Mrbizarro - opened
MiniMax H3 and Pika H3 Flex now run natively on Apple Silicon in Phosphene, using MLX. Thank you, MiniMax and Pika!
The app is free: one installer, no terminal. Pick "H3 Flex" in the top engine picker.
On M4 Max 64 GB, a 1024x576 5 s clip took 16 min with Flex; base H3 at 15 steps took 36.5 min, same shot and seed.
Flex is a preview: sets and side characters can drift in 15 s clips. In your prompt, name everyone who should stay in the scene. Ref2VA isn’t ported yet. Weights · Feedback welcome.
All model credit to MiniMax and Pika. Powered by MiniMax H3.
— Mr Bizarro