OpenAI is starting GPT-6 in ChatGPT across the world. It begins on October 7 for Plus, Pro, Business, and Enterprise. On October 8, it will also show up for Free and Go. The name of the model is not the key part. OpenAI introduced GPT-6 for paying customers in September already.
This October update mainly brings GPT-6 into the core Chat experience for more people. Along with that, OpenAI is adding something called Intelligent UI. This may change how ChatGPT looks and behaves while you use it. Instead of feeling like a basic chatbot, it could feel more like a tool that adjusts the screen as you go. For the main Chat experience, Plus, Pro, Business, and Enterprise will get GPT-6 Sol.
Free and Go will use GPT-6 Luna. There is one exception. The Pro reasoning option still uses GPT-6 Astra, and it does not include Intelligent UI. For Enterprise accounts, what is available can depend on admin settings. This update is only for the usual Chat experience. It does not change what powers ChatGPT Work or Codex. That matters because GPT-6 is already in use in other OpenAI tools and workflows.
Intelligent UI Is the Main Change
In this setup, answers do not have to stay as plain text each time. GPT-6 can show information in other formats when it makes sense. That could mean charts or simple graphics. You may also get tappable buttons, fillable forms, or diagrams you can move through. OpenAI says the system chooses the layout based on what you ask.
For example, if you ask about a bicycle, the reply may include an image, plus parts in it that you can interact with. Intelligent UI will not show up for every message. GPT-6 can still fall back to normal text when that is the better fit.
From One App to Moving Interfaces
That is why Intelligent UI feels like a bigger deal than yet another chat screen update. A lot of older software forces you to work inside one fixed screen. It is built for one task, so you learn that one way. OpenAI is aiming at the opposite. You state what you want, then the system chooses a layout that matches the need.
At times, it may look like a basic calculator for a single question. Other times, it may show a diagram you can interact with. For longer tasks, you could see a visual plan that fits the flow. With this approach, you may not need a new app for each small request. OpenAI also says Intelligent UI uses native building blocks that can stream. It also uses a compiler to assemble the interface as the model talks.
As a result, you see the interface appear in parts. It does not wait for the whole reply to end before anything shows up.
GPT-6 Can Reply Before It Finishes
GPT-6 changes how it builds an answer when a question takes care. It can start writing something you can read before its full behind-the-scenes work is done. After that, it can add more useful points to the same reply. For questions that use web search, OpenAI says it adds a clear, specific claim. In OpenAI tests, GPT-6 Instant began answering web search questions about 44% faster on average than GPT-5.6 Instant.
This points to the time until the first words show up, not the full finish time. OpenAI also notes this is not from an outside, separate benchmark. OpenAI adds that GPT-6 helps with two other areas as well. One is judging more accurately when it should check the web. The other is finding sources that line up with what it says to you.
Final Thoughts
GPT-6 is now moving into the main ChatGPT experience. So more people get to use OpenAI’s latest everyday model. Still, Intelligent UI may be the key shift. With this, ChatGPT feels more like a working tool. You tell it what you want to do. Then the app helps shape how you will end up using the answer.
It is not just plain text. A reply might include a calculator, a shared bill view, or a clear, split-up explanation with visuals. It can also show interactive diagrams. If this trend keeps going, AI helpers could start to look less like simple chat. They may also push back on the small apps that people use for single tasks, not only sites that focus on search.
(Source)