One 3D sailing-ship prompt: Fable 5.2 vs Opus 5.5 vs GPT-6.2 Astra

From the post

I gave this exact prompt to compare Fable 5.5 vs regular Opus 5.5 and GPT-6.2 Astra. Top: Fable 5.5 $ 6.2 Middle: Opus 5.5, $3.9 Bottom: GPT-6.2 Astra $3.5 Opus needed 3 tries to get there, thought for 128,000 tokens twice and hit the output limit.

— @leohuanfang

Author’s claim. Posted before Fable 5.2 came out. The author called it Fable 5.5, the name rumors gave the next Fable. We can’t verify which model ran.