Wan AI is Alibaba's family of video models, and Wan 3.0 is the newest. It turns an image into a lively clip of up to 30 seconds with sound built in. In FacelessGenie, Wan 3.0 is the Pro video tier in Character Video, and you can pick it from the full model list in other formats, so the clip lands inside a finished video with a script, voice, captions and music.
Character Video uses Wan 3.0 on the Pro tier. The cost of every clip shows before you render.
Live credit prices from the video creator.
An example of the Character Video format, where the Pro tier animates each scene with Wan 3.0, with a consistent character, voice and captions.
A short film made with Seedance in the Movie Maker, to compare with Wan's faster, more energetic motion.
Wan 3.0 gives energetic, cinematic motion that stays close to the scene you started from. Action beats, a character running through a market, a camera sweeping over a city, all hold together well. It also generates sound with the clip, and a single clip can run up to 30 seconds, which helps story scenes that need room to play out.
Wan 2.1 and 2.2 were released with open weights, so you can run them yourself on a strong GPU. As of November 2026, Wan 3.0 has no public weights and runs through Alibaba's own app and cloud, and through partners like FacelessGenie. Wan 2.5 and 2.7 are older paid versions. FacelessGenie offers Wan 3.0 for image-to-video because it beats the earlier versions on motion and adds sound, and Wan 2.2 Fast is the default model when a scene is generated straight from text.
The open Wan 2.x models are free to download if you have the hardware to run them. Wan 3.0 is a paid model on Alibaba's platforms as of November 2026, and any free trial there is small and changes often. On FacelessGenie, Wan 3.0 clips use credits from a paid plan, and the cost shows before you render.
Start a Character Video, describe your character and what happens, or paste a script you wrote.
Choose the Pro video tier, or open the full model list and select Wan 3.0. The credit total updates as you switch.
Each scene is animated with sound, then cut together with captions and music. Re-roll any scene in Studio without redoing the video.
Wan is made by Alibaba's Tongyi Lab. The family started as open models and the newest version, Wan 3.0, runs as a hosted model.
Up to 30 seconds per clip in FacelessGenie. Most scenes in a short are a few seconds long, so longer clips matter most for story scenes.
Yes. Wan 3.0 generates sound with the clip. In FacelessGenie you can keep that sound, add a narrator, or both.
Wan 3.0 costs less per clip and moves with more energy. Seedance 2.5 gives the most convincing human acting and lip-sync. Use Wan 3.0 for action and B-roll, and Seedance 2.5 for scenes where people talk to camera.
Videos you make on FacelessGenie are licensed for commercial use. Avoid prompts based on real people, brands or copyrighted characters.
Pick the Pro tier in Character Video and every scene moves with Wan 3.0.