OpenAI is rolling out GPT‑6 in ChatGPT today, extending capabilities first introduced for paid customers to the more than 1.2 billion people who use ChatGPT each week. The release centers on a feature called "Intelligent UI," which lets the model produce responses that combine text with visual and interactive elements — graphics, tappable buttons, forms, charts and other components — directly inside conversations.
What Intelligent UI does
Intelligent UI allows responses to choose the most useful format for the question. A comparison can be presented side‑by‑side, a complex explanation can be shown with an interactive diagram, and a plan (such as cooking timing or travel stops) can appear alongside supporting visuals or maps. The system can also generate ad‑hoc tools inside the conversation, like a savings calculator, a bill splitter, or a simple game.
OpenAI built a library of native, streamable components and a compiler that processes the model‑generated interface so elements can appear progressively while the model generates the response, rather than waiting for the entire answer to complete.
Faster, progressive answers
GPT‑6 can begin answering while it continues to think, interleaving reasoning with partial answers. The model composes responses progressively, adding useful information in each partial response without filler, and keeping the final output cohesive.
In internal evaluations, GPT‑6 Extra High can start answering in the same amount of time as GPT‑5.6 Medium while achieving a better overall score than GPT‑5.6 Extra High. For queries that require web search, GPT‑6 Instant begins answering on average 44% sooner than GPT‑5.6 Instant.
GPT‑6 also improves search‑related decisions and reliability: it more accurately decides when to look something up and finds supporting information more consistently. In internal tests on difficult problems, GPT‑6 addressed the key aspect of users' questions more often than GPT‑5.6.
Safety, alignment and limits
OpenAI says it has updated GPT‑6's safety training based on real‑world use to strengthen protections against high‑risk misuse related to cyberattacks, biological threats and violence. The model builds on several of Astra’s safety advances to better resist attempts to bypass safety training, particularly adaptive, multi‑turn adversarial attacks.
Alignment evaluations were run throughout training to assess honesty and responses to safety restrictions. GPT‑6 is better at recognizing and communicating its own limitations, including when it lacks the information or tools needed to answer.
Availability and model variants
GPT‑6 with Intelligent UI starts rolling out globally today to ChatGPT Plus, Pro, Business and Enterprise tiers in the Chat tab. The rollout expands to Free and Go tiers starting tomorrow; Enterprise availability depends on workplace admin settings.
In ChatGPT, GPT‑6 Sol powers the Plus, Pro, Business and Enterprise tiers, while GPT‑6 Luna powers the Free and Go tiers. Both variants are tuned for everyday conversation. This update affects the Chat experience; the models powering Work and Codex are unchanged in this release.
Why this matters
Intelligent UI represents a shift in how people might interact with conversational AI: rather than forcing every question into the same chat format, ChatGPT can combine interface elements, data and actions to create new ways to explore, understand and act on information. OpenAI frames the release as part of its mission to deliver the benefits of artificial general intelligence safely to as many people as possible, and as a step toward software that adapts to users' needs rather than the other way around.
