OpenAI is rolling out GPT-6 with Intelligent UI in ChatGPT, which lets an answer arrive as charts, tappable buttons, forms and small interactive tools instead of a block of text. Plus, Pro, Business and Enterprise users get it first, and Free and Go users follow the next day. If you have ever asked a chatbot to split a restaurant bill and then done the math yourself anyway, this is aimed squarely at that moment.
TL;DR: ChatGPT Intelligent UI lets GPT-6 build interactive answers, such as calculators, comparisons and forms, inside the Chat tab. OpenAI says it began rolling out on October 7, 2026 for paid plans, with Free and Go next. It does not apply to the Pro reasoning option, and OpenAI admits its design judgment still needs work.
What ChatGPT Intelligent UI actually changes
According to OpenAI’s GPT-6 for everyone announcement, the model now composes a response from text, visuals and interactive elements. Its examples include meal timing for a Sunday roast, wardrobe planning, road trip maps, interactive explanations of the central limit theorem and the Monty Hall problem, and generated tools like a calculator, a bill splitter and a retro game.
The model decides the format on its own. OpenAI’s ChatGPT release notes say it can still reply in plain text when that fits better, so a simple question should not suddenly turn into a dashboard.
Behind the scenes, OpenAI says a native, streamable component library and a compiler render the interface progressively as the model writes it. That is why a chart can start appearing before the full answer is done.
| Plan | Model in Chat | Timing |
|---|---|---|
| Plus, Pro, Business, Enterprise | GPT-6 Sol | Began October 7, 2026 |
| Free and Go | GPT-6 Luna | The following day |
| Pro reasoning option | GPT-6 Astra | No Intelligent UI |
That split also explains why the feature is not a single switch for everyone.
Who gets it and who does not
Enterprise access depends on workspace admin settings. The update covers the Chat tab only, and OpenAI says the models behind Work and Codex are unchanged.
The exclusion most power users will notice is Pro reasoning. That option runs on GPT-6 Astra and does not support Intelligent UI, per the release notes, even though the feature works from Instant through Extra High reasoning on the other models. Anyone paying for the top tier, including the new Pro 500 plan, may find the headline feature sitting just out of reach for their favorite model.
OpenAI does not list iOS, Android or desktop differences. It says only that the experience stays fast and familiar across web and mobile, so check your own app after it updates.
Speed claims and what they leave out
OpenAI says ChatGPT now interleaves thinking with answering, so it can start replying before the reasoning finishes. On questions that need web search, GPT-6 Instant starts answering 44% sooner on average than GPT-5.6 Instant, according to OpenAI’s internal evaluations.
The company also says GPT-6 Extra High begins answering in the same time as GPT-5.6 Medium while scoring better than GPT-5.6 Extra High on everyday agentic tasks. These are OpenAI’s own numbers, and no independent testing is cited.
A faster first word is useful, but it is not the same as a faster correct answer. That is the detail worth watching as more people use it.
What the GPT-6 system card flags
The GPT-6 October system card says Sol and Luna now replace GPT-5.6 Sol and Luna in ChatGPT. It reports instruction hierarchy robustness of 99.99% for Sol and 99.79% for Luna, and gains on factuality across nearly all metrics.
It also lists regressions. Both models slipped on under 18 evaluations for age-restricted content, sexual content and emotional reliance, and OpenAI attributes the emotional reliance result to over-sensitivity to benign nicknames. Its authors judged the production benchmark violations borderline and low severity.
For parents and teachers, those lines matter more than any demo. Interactive answers will feel more authoritative than plain paragraphs, and the safety notes show the models are still being tuned.
How to judge it once it reaches your app
Try it on something you would normally do by hand, like comparing two phone plans or planning a drive. If the result is a form or calculator you actually use, the feature is earning its place. If it is a flashy chart you ignore, plain text was probably the better answer.
OpenAI itself says design judgment still needs improvement, which is an honest caveat. Not a dramatic change, but a visible one, and Free users on Luna will be the first big test of whether it holds up at the scale of more than 1.2 billion weekly users.
