A Year After Opus 4.1 and GPT-5 Wowed Us, Fable/Astra Make Them Look Dated

alejandroll10 · x · 2026-09-05

@SullyOmarr observes that model progress has outpaced public perception: a year ago Opus 4.1, GPT-5 and Gemini 2.5 seemed great, but the new Fable/Astra generation has clearly left them behind. A sentiment piece on the pace of capability jumps; no benchmarks given and the model names remain unverified.

Original post →

More from Models

Models channel →