AI Port Town Benchmark Updated

emollick · x · 2026-07-17

Emad Mostaque announced new results in his benchmark, which tasks AI with **generating an entire port town at once that evolves across historical periods**. The comparison includes models like **GPT-5.6 Pro, Fable, Kimi K3, and Inkling**. He also released an interactive simulation page, noting that these results are "**quite indicative**." The post also points out that Google needs new models to fill the gaps in the current showcase.

Related event: AI Port Town Generation Benchmark Updated(2 posts)→

Original post →

More from Models

Models channel →