@AIWarper On the official GitHub repo of the model (last updated 2 weeks ago) Example #3 (7.mp4) is a similar scene but definitely seems generated / processed.
So... v2v? ¯\_(ツ)_/¯
https://t.co/0yQu9ovvB1
@AIWarper Got suspicious because the detail would have been to good to be true. Gave Claude a screengrab of the video, it identified the location as Dunas de Concón in Chile. "Dunas de Concón dj mix" search on YouTube and here we are.
I should have also taken a look at the repo earlier.
STATION P80X
An animation experiment created with Claude Opus 5.5, running in real time in pure WebGL: no libraries, no 3D models. The whole film is a single 162 KB HTML file, plus my soundtrack as an MP3.
Just released a new ComfyUI node package.
The Variable Text Processor lets you write multiline templates with {variable_name} placeholders and wire any value to named pins. The pin name is the variable name.
Supports STRING, INT & FLOAT. Available on the Comfy Registry.
https://t.co/z3lfyi03Tc
#ComfyUI
Seriously impressed by Z-Image from @Ali_TongyiLab.
Pushing image fidelity with a Z-Image ComfyUI pipeline distilled from years of tinkering.
Huge props to the team.
#AIart#GenerativeArt#LatentSpace