OpenAI's biggest model yet is now in your chat window. On October 7, 2026, OpenAI began rolling out GPT-6 across ChatGPT — starting with paid subscribers on GPT-6 Sol, and extending to Free and Go users on GPT-6 Luna the following day.
The headline feature is Intelligent UI — a rethinking of how ChatGPT responds. Instead of walls of text, the model can now generate charts, buttons, forms, diagrams, maps, and small interactive tools directly inside a conversation. Ask it to build a savings calculator or a bill splitter, and it does — right there in the chat window. 9to5Mac's breakdown of the Intelligent UI captures just how different the experience feels compared to GPT-5.6.
For practical tasks, it's a meaningful shift. The interface no longer just tells you; it shows you, and lets you interact with the output. A question about budgeting doesn't return a bulleted list — it returns a working calculator.
On the performance side, OpenAI says GPT-6 Instant starts answering web-search questions about 44% sooner than GPT-5.6 Instant in internal testing. That gap translates to noticeably snappier responses on factual queries — a detail that matters most for the free-tier users getting GPT-6 Luna.
There are limits to the rollout. The most capable variant — GPT-6 Astra, which runs on Pro-tier ChatGPT — is not part of this wave. The older macOS and Windows desktop apps are also excluded for now, and Enterprise workspace access depends on whether the administrator has enabled it. The Dots product, which introduces always-on agents powered by GPT-6 Astra, remains limited to Pro and Business Premium customers.
The model arriving for free users is GPT-6 Luna, a lighter version tuned for speed and general use. Paid tiers (Plus, Pro, Business) get GPT-6 Sol, the more capable mid-tier variant. The gap between what free and paid users get has always existed — with this rollout, the baseline for everyone is substantially higher than it was a week ago.
What stands out most is the scope: making GPT-6 available to all ChatGPT users simultaneously (or within 24 hours) is a deliberate statement. OpenAI has made the case before that its best models should be accessible without a paywall. This rollout follows that logic more aggressively than any previous release.