What answers look like now
On October 7 OpenAI published "GPT-6 and Intelligent UI for everyone." The core change is Intelligent UI: GPT-6 has been trained to combine text, visuals and interactive elements into one answer depending on the question. OpenAI's examples include diagrams when asked how an airplane wing generates lift, a map marking each stop of a multi-day hike, and a savings calculator or bill splitter built on the spot when you need to do the math. When plain text is most useful, it still gives just text.
According to OpenAI, it is powered by a library of native, streamable components plus a compiler that draws the interface progressively as the model generates. Users can turn visuals down like other personalization settings. OpenAI product manager Aarush Selvan told TechCrunch that ChatGPT "has predominantly been a text-based interface," while the most helpful answers "aren't just text."
Who gets which model
Paid and free users get different models. Plus, Pro, Business and Enterprise get it globally from October 7, running on GPT-6 Sol; Free and Go get it from October 8, running on GPT-6 Luna. OpenAI says both are tuned for everyday conversation. The models behind Work and Codex are unchanged this time, and Enterprise availability depends on admin settings.
What it means
For 1.2 billion weekly users, this is a clear shift in ChatGPT's interface: Q&A moves from reading text to using small tools, and the model decides when to draw without users having to ask. OpenAI itself admits the model's "design judgment" needs work, and the numbers inside charts and calculators are model-generated too, so for decisions about money or health, check the underlying data. Developers making content or tutorials around ChatGPT should expect more of these visual answers to show up in users' screenshots.
via: OpenAI announcement, TechCrunch report, FourWeekMBA report