What's this about?
OpenAI updated what happens inside ChatGPT so that its stronger "GPT-5.6 Sol" model now runs both the fast "Instant" replies and the slower "deep reasoning" answers for Plus and Pro users, and it made the better models easier to reach across free/Plus/Pro tiers. It's a plumbing change to the product, not a brand-new model launch.
What exactly is "Instant vs. deep reasoning"?
Modern chat assistants secretly pick between two speeds. "Instant" is a quick, cheap answer for easy questions; "deep reasoning" lets the model take longer and think step-by-step for hard ones. Think of a colleague who fires back a one-liner for a simple question but goes away and drafts a memo for a hard one. The news is that the *smarter* model now drives both speeds, so even the quick answers get better.
What's the community saying?
14.5k likes on the announcement; devs already probing the details.
- Practical confusion: simonw publicly asked what the API equivalent of "GPT-5.6 Instant" even is — the consumer names don't map cleanly to API model IDs.
- Access-tier questions: which capabilities land on free vs. Plus vs. Pro, and whether limits changed.
- Framing debate: some see "better model for everyone" as the real competitive move (make the good model the default) rather than chasing a new headline model.