Any
$500
5
Oct 3, 2026
We’re looking for an experienced AI engineer to build a self-hosted GPU environment for advanced AI video and image editing.
Our main goals are:
Video character swapping / replacement
Video face swapping
Image face swapping
Character animation from a reference image + driving video
Strong identity preservation
Maximum control over the models and workflows
Models we’re currently looking at include:
Wan2.2-Animate-2-14B
Wan2.2-Animate-14B
Wan2.1-VACE-14B
We’re also open to better models or combinations if you recommend them.
The final setup should ideally run through ComfyUI or another simple interface, hosted on something like RunPod, Vast.ai or a similar GPU provider.
To apply
We don’t want a generic application.
Tell us how you would actually build this setup:
Which models/tools would you combine?
Are Wan Animate and VACE the right choices?
What would you use for face swap vs full character replacement?
Which GPU / VRAM would you recommend?
What quality level is realistically possible today?
What are the main limitations?
If possible, send a 2–5 minute Loom/video explaining your approach.
We care more about your technical opinion and understanding of the project than a long application.
This is a one-time project, with potential additional AI projects afterward.