GPT-6 with Intelligent UI rolls out to all 1.2 billion ChatGPT users

OpenAI has started pushing GPT-6 into the main Chat tab for all ChatGPT users, replacing GPT-5.6 Sol and Luna across every plan tier. The rollout began on 7 October 2026 for Plus, Pro, Business, and Enterprise accounts, with Free and Go users following on 8 October 2026. That is more than 1.2 billion weekly users receiving a new default model in the space of two days.

The update also introduces something new alongside the model upgrade: Intelligent UI, a capability that lets ChatGPT respond with interactive elements embedded directly in the reply rather than plain text.

What changed and for whom

GPT-6 Sol is now the default model for paid ChatGPT tiers (Plus, Pro, Business, and Enterprise). GPT-6 Luna takes over for Free and Go users. Both replace their GPT-5.6 equivalents in the Chat tab.

You do not need to change any settings. The new model appears automatically as the rollout reaches your account. If you want to compare outputs against the previous generation, GPT-5.6 Sol is still available in the model picker. Enterprise accounts may see a delay depending on how workplace admins have configured their deployment settings.

One thing worth noting: this update applies to the Chat tab only. The models powering ChatGPT Work and Codex are not changing as part of this release.

What Intelligent UI actually is

Intelligent UI is not a new model. It is a response format that sits on top of GPT-6, giving ChatGPT the ability to produce interactive elements as part of an answer rather than always defaulting to prose.

OpenAI describes it as “a library of native, streamable components, along with a compiler that processes the interface as the model generates it.” In practice, that means a reply can include a chart, a form, a clickable calculator, a bill splitter, or even a simple game, and those elements appear progressively as the model generates them rather than loading all at once at the end.

The practical examples OpenAI gives are fairly grounded: a cooking timeline appearing beside a recipe, road-trip stops plotted on a map, a savings calculator built on demand, or a learning exercise where a user can adjust variables and see results change in real time. The model decides whether a visual or interactive element genuinely improves the answer, or whether a plain text response is the right call.

OpenAI frames this as “a step toward a ChatGPT that can shape an interface around what you’re trying to do,” which is a useful way to think about where this is heading even if the October 2026 version is still early.

Faster answers, progressively delivered

Alongside Intelligent UI, GPT-6 introduces what OpenAI calls streaming thinking. Rather than completing the entire reasoning process before showing any output, GPT-6 begins answering while it is still thinking, adding information progressively. OpenAI says this cuts average wait times for live web search answers by 44 per cent compared with GPT-5.6 Instant.

For most conversations that is a noticeable quality-of-life improvement, particularly for queries that involve research or multi-step reasoning.

Where GPT-6 fits in the wider model family

GPT-6 Sol and Luna were introduced in late September 2026 as more affordable successors to GPT-5.6, with API pricing around 50 per cent lower. Sol is priced at $2.00 per million input tokens and $10.00 per million output tokens; Luna sits at $0.10 in and $0.50 out.

At the top of the GPT-6 family sits GPT-6 Astra, which OpenAI describes as its most capable model. Sol and Luna are smaller models trained with methods similar to Astra’s, designed to carry some of its strengths, particularly in professional and reasoning tasks, at a lower cost. This ChatGPT rollout is about bringing those two models to a consumer-scale audience as the new default.

What this means for you

For most ChatGPT users, the change is largely invisible on day one. You open the Chat tab, and GPT-6 is there. The responses may feel faster, and you will start to see interactive elements appear where ChatGPT judges them genuinely useful, though not every reply will look different.

The more meaningful shift is structural. Intelligent UI changes the assumption that a chat interface should always produce text. A question about splitting a dinner bill might now produce a working calculator. A question about a financial concept might produce an adjustable chart rather than three paragraphs. Whether that is an improvement depends on the context, and OpenAI has said the model is trained to make that judgement rather than forcing interactivity on every answer.

For business and enterprise teams, the upgrade lands in the Chat tab with no action required, though admins should check their deployment settings if they want to manage timing or availability across their organisation.

For developers, the GPT-6 Sol and Luna models themselves have been available via the API since their introduction on 22 September 2026. Intelligent UI is a ChatGPT-specific response format and is not part of the standard API at this stage.