OpenAI began expanding GPT-6 access in ChatGPT on October 7, adding interactive interfaces and progressively delivered answers across eligible subscription tiers.
OpenAI began rolling out GPT-6 in ChatGPT on October 7, 2026, extending access beyond paid customers who received the first GPT-6 models the previous month. The release adds a feature OpenAI calls Intelligent UI, which can present answers using interactive elements alongside text. The company says examples include visual explanations, charts, forms, maps, and small tools such as calculators or bill splitters.
The interface is generated from a library of components, with a compiler that lets parts of a response appear while the model is still producing it. OpenAI says GPT-6 chooses whether an interactive layout fits the question; straightforward prompts can still receive text responses. The feature is designed for ChatGPT’s web and mobile experiences.
The update also changes how answers are delivered: ChatGPT can begin responding while GPT-6 continues working, adding information in successive portions. OpenAI reports that GPT-6 Instant starts responses to questions requiring web search 44% sooner on average than GPT-5.6 Instant. In an internal evaluation of everyday agentic tasks, the company says GPT-6 Extra High began answering in the same time as GPT-5.6 Medium while scoring better overall than GPT-5.6 Extra High. These are company-reported evaluation results, not independent benchmark findings.
OpenAI says GPT-6 with Intelligent UI started rolling out first to Plus, Pro, Business, and Enterprise subscribers, with Free and Go access scheduled to begin October 8. Enterprise users’ access depends on their administrators’ settings. The ChatGPT experience uses GPT-6 Sol for the paid tiers and GPT-6 Luna for Free and Go; OpenAI says the models used in Work and Codex are unchanged by this release.
The announcement also describes safety training changes targeting high-risk misuse and improved recognition of model limitations. OpenAI says the work builds on safety advances from Astra and is intended to strengthen resistance to adaptive, multi-turn attempts to bypass safeguards.