OpenAI has begun rolling out its GPT-6 models across ChatGPT, bringing GPT-6 Sol to paid subscribers and GPT-6 Luna to free users — and introducing a new capability the company calls Intelligent UI, which turns chatbot answers into dynamic, interactive interfaces. According to 9to5Mac's report on the announcement, the rollout began Thursday for free tiers, with paid users gaining access first.

Which Users Get Which Model For more context on this story, see our ongoing latest AI developments.

GPT-6 Sol is rolling out now in ChatGPT for Plus, Pro, Business, and Enterprise subscribers. GPT-6 Luna will reach Free and Go users starting Thursday. OpenAI says both models are tuned for everyday conversation — a signal that the company is optimizing its consumer flagship for general-purpose use rather than reserving it for specialized workloads.

The staged rollout gives OpenAI's paying customers first access to the more capable variant while the free tier receives Luna, keeping ChatGPT's massive free user base on a current-generation model. The move comes as OpenAI's models have already rolled out through the API; this step puts them directly in front of ChatGPT's consumer audience.

The scale of that audience is what makes the rollout significant. OpenAI has previously reported that ChatGPT is approaching one billion weekly active users, a figure the company has cited as it prepares for a potential public listing. Moving that entire user base onto GPT-6-class models in a matter of days — rather than over months of gradual migration — is an aggressive cadence, and one that will test the company's inference infrastructure at a scale no AI product has operated at before.

What Intelligent UI Actually Changes

The headline feature is Intelligent UI, a new response format that lets ChatGPT dynamically combine text, visuals, and interactive elements based on what a question requires. Instead of a static block of text, answers can now include graphics, tappable buttons, forms, and charts, rendered inline within the conversation.

The format adapts to the query, according to OpenAI's demonstration. A comparison question might be answered side by side. An explanation might be clearer as an interactive diagram. And when plain text is genuinely the best answer, ChatGPT can still respond with one.

For more complex prompts, ChatGPT can generate interactive widgets that let users explore concepts or adjust inputs to see how results change — a step beyond static images, and closer to a lightweight application environment living inside the chat window.

Mini Apps Inside the Chat

OpenAI showcased ChatGPT building interactive tools directly within a conversation: calculators, bill splitters, and games, all generated on demand. That capability blurs the line between chatting and using software. Rather than routing users to a separate tool or app, the assistant assembles the tool in place — and discards it when the task is done.

If the approach holds up in practice, it could reshape expectations for what a chatbot interface is. The conversation becomes a surface where bespoke, single-purpose micro-apps appear, serve their purpose, and disappear — no install, no menu, no separate workflow.

Answers That Begin Before Reasoning Ends

Alongside the visual changes, GPT-6 models in ChatGPT introduce a change to how responses unfold. OpenAI says ChatGPT can now begin answering while it continues to think, progressively adding information to the response rather than waiting for the entire reasoning process to finish.

For users, that means less waiting at the start of a response and a more incremental flow of information — closer to how a knowledgeable colleague might talk through a problem, refining the answer as they go. It also masks latency on harder prompts, where reasoning models traditionally leave users staring at a thinking indicator.

A Crowded, Price-Cutting Market

The rollout lands in the middle of an intensifying price war among frontier AI providers. The Wall Street Journal reported this week that the competition on price is heating up, with OpenAI gaining ground on rival Anthropic. Anthropic, for its part, launched Claude Haiku 5.5 this week with a steep price cut on its small model, a move widely read as a direct challenge to OpenAI's cheaper tiers.

The specifics underline how fast the floor is dropping: MarkTechPost reported that Claude Haiku 5.5 carries a 1 million-token context window at $0.10 per million input tokens — pricing that puts long-context AI within reach of high-volume, cost-sensitive applications. When a frontier lab's small model offers a million tokens of context for a dime, "affordable AI" stops being a marketing line and becomes a commodity expectation.

Against that backdrop, Intelligent UI is as much a strategic play as a feature release. Model quality alone is no longer a durable differentiator when rivals ship competitive models within days of each other. Interaction design — how answers look, feel, and behave — is the surface OpenAI is betting on to keep ChatGPT's billion-scale user base engaged.

What to Watch

The immediate question is rollout stability: how quickly Luna reaches all free users, and how Sol performs under full production load. The longer-term question is whether interactive answers become a new default expectation for AI assistants — and whether OpenAI's rivals follow with interface innovations of their own, or concede the user-experience high ground. Either way, the chatbot's answer format just became a competitive battleground of its own.

---

Stay Ahead of AI

Get the latest AI news, analysis, and breakthroughs — all in one place.

Read more AI news →