Wan 2.2: AI video on your own machine
Generate video from text or an image — Apache 2.0
- Video memory
- 16 GB
- Setup
- ~30 min
- Access
- port 8188
- Billing
- hourly, in BRL
Generative video is where the GPU matters most — and where platforms charge the most per credit. Wan 2.2 in its 5-billion-parameter variant fits a 16–24 GB card, does text-to-video and image-to-video, and is Apache 2.0.
Video generation with Wan 2.2 (5B variant), fitting 16–24 GB cards. Text-to-video and image-to-video, with model, text encoder and VAE preloaded and ComfyUI running. Apache 2.0: commercial use allowed.
What it is for
- Short ad clips from a line of text or a product photo
- Animate a still catalog image for posts and stories
- Test dozens of variations without burning per-generation credits
- Commercial use allowed, with no revenue ceiling like rival models
How to deploy
- Create your account and add balance (card or Pix, no subscription).
- In the console, pick the Wan 2.2 (vídeo) template and a machine — the console hides the ones that do not meet the requirement.
- In about 30 minutes the setup finishes and the access address shows up in the panel, on port 8188.
Done? Just destroy the machine and billing stops with it. No contract, no minimum commitment.
FAQ
How long does a clip take?
Two to ten minutes, depending on card, resolution and length. A bigger card shortens it a lot — which is where renting by the hour pays off.
Which card should I pick?
16 GB is the minimum. 24 GB means longer clips and shorter waits.
Can I use it commercially?
Yes, it is Apache 2.0. Some open competitors only allow commercial use below a revenue threshold; this one does not.