OpenAI quietly changed how GPT-5.6 Sol is positioned in Codex.
Its description went from “Latest frontier agentic coding model” to “Reliable agentic workhorse for everyday tasks,” while its catalog priority moved from 6 → 4.
Could just be catalog cleanup, but this is exactly the kind of repositioning you’d expect before the next model lineup arrives.
(In my view, this is an explicit preparation to replace GPT-5.6 Sol with GPT-6 Sol.)