Computational Model Studio · Model release

Wan2.1 T2V 1.3B

Generates short video from text prompts using a downloadable 1.3B checkpoint.

Developer: Wan Team · Last verified: 2026-09-30

What it is designed to do

Generate video from a text prompt at the documented 480p setting.

Documented uses: Text-prompted 480p video generation.

Wan2.1 T2V 1.3B model files — available. Public publisher repository; review applicable terms before use. Artifact documentation
- workstation local: NVIDIA RTX 4090 GPU; about 8.19 GB GPU VRAM; Publisher model requirement for text-to-video at supported 480p. The RTX 4090 is the publisher's single-GPU example with model and T5 CPU offload; the source does not specify that the 8.19 GB requirement covers every prompt or output length. (publisher stated requirement) Limits: Performance varies by runtime configuration and hardware; This does not establish laptop or mobile suitability.
The full artifact-specific rights document has not been separately reviewed; consult the publisher repository conditions.

Technical profile and review notes

Inclusion rationale: Adds video output and a publisher-conditioned workstation-local deployment example.

Limitations

The publisher recommends 480p for this checkpoint; 720p results are less stable and the provided run matrix does not support that mode.
Family-level image-to-video or video-to-audio claims are not transferred to this text-to-video checkpoint.