Cmd
NVIDIA🇺🇸 USMajor lab· Aug 13, 2026
Videoopen weights
Research checkpoints for causal image-to-video generation that produce videos autoregressively in chunks. Variants cover short and long clips plus camera control, built on Cosmos-Predict2.5 2B.
Summary of the model card
- License
- other
- Modalities
- text, image → video
1 variant
| Standard | HF |
Sources
HuggingFacePaperDeutsch
Forschungs-Checkpoints für kausale Bild-zu-Video-Generierung, die Videos blockweise autoregressiv erzeugen. Varianten für kurze und lange Clips sowie mit Kamerasteuerung, Basis ist Cosmos-Predict2.5 2B.
Provenance
- Detected
- Sep 27, 2026
- First source
- huggingface
- Description
- Summary of the model card
- Editorially reviewed
- —