Wan 3.0 by fal
Create videos up to 30 seconds from text, start and end frames, or mixed image, video, and audio references—with native sound.
Best for
- text-to-video
- start/end frames
- multimodal references
- native audio
Use when
- You want one continuous shot from 2 to 30 seconds.
- You want to anchor motion with a first frame and optional last frame.
- You need up to 10 image, 5 video, and 5 audio references in one request.
- You want native sound generated with the video.
Avoid when
- You need to continue an existing clip; use Wan 2.7 in Extend.
- You need Frames and rich References in the same request; choose one input mode.
Inputs
- prompt
- optional first and last frames
- up to 10 reference images
Outputs
- video with optional native audio
DayGen surfaces
- /video/create
- /video/direct
- Agent video generation
Available settings
- Mode
- Text, Frames, References
- Aspect Ratio
- Adaptive, 16:9 (Wide), 9:16 (Story), 1:1 (Square), 4:3 (Classic), 3:4 (Portrait). The model chooses one fixed output shape from your prompt, frames, and references. The result is not responsive.
- Resolution
- 480p (Faster), 720p, 1080p (Higher Quality)
- Duration
- Auto, 2–30s
- Audio
- On, Off, Up to 5 audio references
- Advanced settings
- Prompt expansion, Document URL, Web page URL