Image-to-Video
Diffusers
Safetensors
MLX
cosmos3
video-generation
quantization
int4
w4a16
sdnq
apple-silicon
cuda
8-bit precision
Instructions to use JuliaML/Cosmos3-Super-Image2Video-4Step-INT4-G64-BF16 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use JuliaML/Cosmos3-Super-Image2Video-4Step-INT4-G64-BF16 with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline from diffusers.utils import load_image, export_to_video # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("JuliaML/Cosmos3-Super-Image2Video-4Step-INT4-G64-BF16", dtype=torch.bfloat16, device_map="cuda") pipe.to("cuda") prompt = "A man with short gray hair plays a red electric guitar." image = load_image( "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/guitar-man.png" ) output = pipe(image=image, prompt=prompt).frames[0] export_to_video(output, "output.mp4") - MLX
How to use JuliaML/Cosmos3-Super-Image2Video-4Step-INT4-G64-BF16 with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] huggingface-cli download --local-dir Cosmos3-Super-Image2Video-4Step-INT4-G64-BF16 JuliaML/Cosmos3-Super-Image2Video-4Step-INT4-G64-BF16
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- LM Studio
- Atomic Chat
| { | |
| "format": "cosmos3-sdnq-uniform-community-v1", | |
| "status": "complete", | |
| "started_at": "2026-08-11T10:18:28+00:00", | |
| "source": "models/Cosmos3-Super-Image2Video-4Step/transformer", | |
| "output": "checkpoints/Cosmos3-Super-Image2Video-4Step-SDNQ-INT4-G64-BF16IO/transformer", | |
| "weights_dtype": "int4", | |
| "group_size": 64, | |
| "activation_dtype": "bfloat16", | |
| "bf16_io": [ | |
| "proj_in", | |
| "proj_out" | |
| ], | |
| "use_svd": false, | |
| "use_hadamard": false, | |
| "use_dynamic_quantization": false, | |
| "use_quantized_matmul_by_default": false, | |
| "quantization_device": "mps", | |
| "quantization_seconds": 84.7199640829931, | |
| "quantized_matrix_count": 896, | |
| "parameter_bytes": 38334765440, | |
| "finished_at": "2026-08-11T10:20:57+00:00", | |
| "save_seconds": 63.97164558299119, | |
| "total_seconds": 148.69399683299707, | |
| "safetensor_bytes": 38335018664, | |
| "safetensor_gib": 35.7022682800889, | |
| "sdnq_version": "0.2.4" | |
| } | |