GPT-6 vs Fable 5.1 on Code4Scene: floating objects sink Fable's 3D scenes

Lianhuiq · x · 2026-09-11

The Code4Scene benchmark compares GPT-6 and Fable 5.1 on identical prompts and asset packs, both generating a harbor scene with cranes, containers, and vehicles. GPT-6 gets placement mostly right (still imperfect), while Fable repeatedly leaves objects floating. The benchmark stresses spatial details — containers must stack without gaps, wheels touch the road, trucks need clearance.

Related event: GPT-6 vs Fable 5.1 Tested on 3D Harbor Scene Generation(2 posts)→

Original post →

More from Models

Models channel →