What happened
OpenAI began a global rollout of GPT-6 in ChatGPT on October 7, 2026, together with a new response format it calls Intelligent UI. Paid plans received it first; Free and Go users followed from October 8. The update replaces the GPT-5.6 models previously used in the Chat tab.
What’s new
| Plan | Model | Rollout start |
|---|---|---|
| Plus, Pro, Business, Enterprise | GPT-6 Sol | October 7, 2026 |
| Free, Go | GPT-6 Luna | October 8, 2026 |
- Intelligent UI: instead of returning only text, the model composes answers from text, visuals and interactive elements such as charts, buttons, forms and small tools (OpenAI’s examples include a calculator, a bill splitter and a game). It chooses the layout per question and falls back to plain text when that works best.
- Streaming while reasoning: GPT-6 can start answering before it finishes reasoning, and the interface renders progressively using a native component library and compiler.
- Latency: OpenAI reports that GPT-6 Instant starts answering web-search questions 44% sooner on average than GPT-5.6 Instant. This measures time to first answer, not total completion time.
- Reasoning range: according to OpenAI’s release notes, Intelligent UI works from Instant through Extra High reasoning; the Pro reasoning option still uses GPT-6 Astra and does not support Intelligent UI.
OpenAI’s announcement does not state usage limits. Enterprise access depends on workspace admin settings.
Why it matters
This is the default model change for ChatGPT’s entire user base, so it affects anyone who relies on ChatGPT output in daily work: answers may now arrive as interactive components rather than copyable text. Teams that document ChatGPT-based workflows or train staff on them should check that prompts still produce the format they expect.
For developers, the consumer rollout is separate from the API. The model names used here (gpt-6-sol and gpt-6-luna) are also API model IDs, and OpenAI separately shipped GPT-6.1 Sol and an “Ultrafast” mode for the API and Codex in the same week, priced at $12 per million input tokens and $60 per million output tokens for Ultrafast.
How to try it
- Open ChatGPT on any plan; the new model is the default in the Chat tab once the rollout reaches your account.
- Ask for something that benefits from interaction, such as a loan comparison or a unit converter, to see Intelligent UI.
- If you need plain text for copying into documents, say so in the prompt.