GPT-6 Intelligent UI can show charts, diagrams, buttons, forms, and tools that work directly inside a ChatGPT conversation when the system judges that format useful. No particular interface is guaranteed for any given request. You can choose a simpler style in the web settings to reduce visuals, but this does not force every reply into text only. The capability belongs to Chat, not the Work tab or the Pro reasoning mode.
When do charts, comparisons, and interactive tools appear?
OpenAI announced Intelligent UI on October 7, 2026. Instead of always producing a text block, ChatGPT can choose a side-by-side comparison, an interactive diagram, a chart, a map, or a tool that helps complete a task within the conversation. Official examples include calculators, bill splitters, and games. A straightforward question can still receive a text-only answer.
The format depends on the task and context, not on a magic word in the prompt. Users can explicitly request a calculator or a shorter reply, but the documentation does not promise a specific component on every attempt. An answer may also begin appearing while the model continues reasoning or consulting tools, with further details arriving later.
What changes in how a response is assembled?
According to OpenAI’s technical explanation, GPT-6 composes responses from a library of visual and interactive components. A processing system lets these pieces appear progressively rather than waiting for the entire response to be completed. This is the vendor’s description, not an independent performance measurement.
There is a practical difference between an illustrative chart and an interactive control: the latter lets readers change values, select options, or explore parts of an explanation inside the conversation. The presence of controls does not prove that calculations, underlying data, or conclusions are correct; important information still requires verification.
How can you reduce visuals in ChatGPT?
On the ChatGPT website, the dedicated FAQ says to open Settings, go to Personalization, find Layout and Visuals, and select Simple. A separate OpenAI help page instead describes turning off Layout and visuals. Because the two official instructions use different labels, look for the appearance control actually shown in your account rather than assuming every interface has the same button.
This reduces visual and interactive elements, but it does not completely disable Intelligent UI or guarantee a return to a classic or text-only experience. You can also ask for a concise answer in paragraphs with fewer graphics within a conversation, or save the preference in Custom Instructions. Such preferences influence presentation without determining the exact final layout.
Which plans, modes, and channels support it?
The October 7 announcement began a global rollout in Chat for Plus, Pro, Business, and Enterprise, with expansion to Free and Go indicated from October 8. Beginning a rollout does not mean every account receives the feature at once. Enterprise administrators can manage access. The FAQ links Intelligent UI to GPT-6 Sol on the named paid plans and GPT-6 Luna for Free and Go, subject to eligibility and gradual availability.
Model documentation also says Free and Go accounts may still use GPT-5.6 Luna as their default model. It is therefore not sound to assume every user on those plans already sees GPT-6. Available reasoning levels and model-picker options also depend on the plan and permissions.
| Environment or mode | Intelligent UI status |
|---|---|
| Chat with GPT-6, Instant through Extra High | Supported when the account has access; the response format can still vary. |
| Chat, Pro reasoning (GPT-6 Astra) | Intelligent UI is not available at this level. |
| Work tab | Intelligent UI is unavailable; Work is a separate environment for tasks. |
| Voice | The FAQ says Intelligent UI does not work in Voice mode. |
| Chat on the web and in updated supported apps | Gradual rollout. Older macOS and Windows desktop apps do not support it; the web is the stated alternative. |
What does this customization not control?
Intelligent UI has no separate usage quota. The model, tool, and plan limits already in place still apply. Changing the response style does not enable GPT-6 for an account without access, change how Work operates, activate the feature in Voice, or guarantee a game, map, or form on every reply.
The FAQ also says that some controls, such as checklists, may retain their state after refreshing the same conversation, but state is not carried across conversations. Connected plugins continue to work in ChatGPT; this does not mean every integration will display a dedicated interactive interface.
What should you check if it does not appear?
Check that you are in Chat, that GPT-6 and an Instant-through-Extra-High reasoning level are available to your account, and that you are using the website or an updated compatible mobile app. In a managed workspace, administrator policy can also restrict access. The absence of visuals in one reply does not establish a malfunction: ChatGPT may choose plain text because it is sufficient for the question.
This explanation draws on the announcement, release notes, and help pages consulted on October 25, 2026, along with Canaltech’s reporting. It is not an operational test of the interface, verification of any individual account, or a promise of a fixed format.
Sources and references
- OpenAI, October 7, 2026 Intelligent UI announcement
- OpenAI, Intelligent UI FAQ
- OpenAI, GPT-6 and other models in ChatGPT (English documentation)
- OpenAI, GPT-6 and other models in ChatGPT (Portuguese translation)
- OpenAI, ChatGPT release notes for October 7, 2026
- Canaltech, Viviane França’s October 8, 2026 report
