A developer has released "Three.js Object Sculptor," an open-source Codex plugin that takes a single photo of any object and, using GPT-5.6 Sol and Codex, generates an editable Three.js 3D model written entirely in code. Published on GitHub around July 9, 2026, it drew attention with a case in which a reference photo of a Chinese-style ship produced a 3D model whose silhouette and main structure were recognizable.
Continue reading
The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in✓ Signed in — this article isn’t included in your current plan.