Cosmos3-Super-Text2Image-4Step
NVIDIA🇺🇸 USMajor lab· Jul 7, 2026
Imageopen weights
Text-to-image model from NVIDIA, distilled to 4 sampling steps via DMD2 without classifier-free guidance. The modelcard reports up to 25x fewer diffusion evaluations per image than the 50-step base model with CFG.
Summary of the model card
- License
- other
- Modalities
- text → image
1 variant
| Standard | HF |
Sources
HuggingFacePaperDeutsch
Text-zu-Bild-Modell von NVIDIA, per DMD2 auf 4 Sampling-Schritte destilliert und ohne Classifier-free Guidance. Laut Modelcard sind dafuer bis zu 25-mal weniger Modellauswertungen pro Bild noetig als beim Basismodell mit 50 Schritten.
Provenance
- Detected
- Sep 27, 2026
- First source
- huggingface
- Description
- Summary of the model card
- Editorially reviewed
- —