Wan 3.0 Is Live on Aristotto

Wan 3.0 is live: native 30-second video with audio, keyframes, and AI Director guiding your shots. Now included in our 46% off credits promo through Dec 20.

Wan 3.0 is Alibaba's latest video model, and its biggest shift is scope: a single generation now runs up to 30 seconds at up to 1080p, with smart duration that matches clip length to your prompt instead of padding a short scene out. It reads image, video, and audio references together, and supports first-and-last-frame keyframes so you can define exactly where a shot starts and ends. AI Director reads your shot language directly, so naming moves like "slow orbit right" or "handheld tracking shot" gets you a more consistent camera path than describing the scene alone. Audio works differently here than on most models: it's generated from what you describe in the prompt, not uploaded. Write dialogue as the exact line you want spoken, and name ambient sound and action cues explicitly rather than leaving them implied. Wan 3.0 also holds up well on human faces, keeping natural micro-expressions consistent across a shot. Promo: Through December 20, generating with Wan 3.0 costs 46% fewer credits, automatically, on every paid plan. Your subscription price stays the same. Not sure what to test? A woman stands at a rain-streaked bus stop at night, neon reflections rippling on the wet pavement. She glances at her phone, exhales slowly, then looks up as headlights sweep across her face. She says, "I didn't think you'd actually come." Camera holds a slow push-in throughout. Rain patters on the shelter roof, distant traffic hum, her voice quiet and steady over the ambient noise.