An AI product-visualization experiment
Launch imagery for objects that ask to be read closely.
ARCTYK shows what current AI models can do in product visualization: CGI stills, motion, material variants, and web-ready 3D for furniture and consumer products.
Every model, material, and render here was built in Blender through code written with a language model: an experiment in how far an AI-driven 3D workflow can go.
Capabilities
- Still imageryLaunch frames, catalogue elevations, material details, and lifestyle applications.
- Motion and variantsTurntables, orbit studies, and coordinated color or finish families from one product asset.
- Interactive 3DBrowser-ready GLB preparation for dimensional review, ecommerce, and product storytelling.
Visual index · 02 dossiers
Selected dossiers, sequenced for a close read.
VELA — plate 01ivory shell / smoked ash base Furniture launch visualization
VELA
2026
An upholstered lounge chair developed into a reusable launch system spanning catalogue views, construction studies, color variants, lifestyle imagery, motion, and browser-ready 3D.
NULLE — plate 02graphite hub / power hierarchy Consumer electronics visualization
NULLE
2026
A five-port desktop USB-C power hub visualized as a coherent launch system across stills, motion, and browser-ready 3D.
How this was made
Curious about the workflow?
Every plate here started as code a language model wrote, directed and refined in Blender. Questions about the workflow are welcome, and larger projects that need it are considered case by case.
- ModelingGeometry generated from code rather than sculpted by hand.
- VariantsNew colors and finishes from one model, at almost no extra cost.
- OutputStills, motion, and browser-ready interactive 3D.