Alibaba's newest video generation model, Wan 3.0, is now available in the video model picker — with smoother motion, better scene fidelity, native audio, and clips up to 30 seconds.
Wan 3.0
Wan 3.0 (wan-3.0) covers both text-to-video and image-to-video in a single pick: add a first frame to animate a still, and optionally a last frame to control where the clip lands.
- 2–30 second clips at 480p, 720p, or 1080p (1080p by default)
- Native generated audio, on by default
- Aspect ratios 16:9, 9:16, 4:3, 3:4, and 1:1
- Pricing: $0.065/s at 480p, $0.13/s at 720p, $0.26/s at 1080p
Wan 3.0 Prime
Wan 3.0 Prime (wan-3.0-prime) is the higher-fidelity image-to-video SKU — same controls, more cinematic motion and visual continuity from your source image.
- Image-to-video only (first frame required, optional last frame)
- Pricing: $0.0884/s at 480p, $0.182/s at 720p, $0.364/s at 1080p
Wan 3.0 R2V
Wan 3.0 R2V (wan-3.0-r2v) generates from up to 10 reference images, keeping subjects and style consistent while following your prompt. Cite references positionally in the prompt ("the subject in Image 1 walks past the storefront").
- Same duration, resolution, and audio options as Wan 3.0
- Same per-second pricing as Wan 3.0 — reference images carry no surcharge
Wan 2.6, Wan 2.6 Flash, and Wan 2.7 R2V remain available and unchanged.