# GPT-6 reaches all ChatGPT users with interactive "Intelligent UI" as OpenAI adds 512 MB audio uploads for paid plans

Free and Go users receive GPT-6 Luna starting today, and OpenAI says GPT-6 Instant starts answering web-search queries 44% sooner than GPT-5.6 Instant, a figure that measures time to first token and not total answer time.

- Published: 2026-10-08T06:25:29.519Z
- Canonical: https://polylog.news/ai/2026-10-08/gpt-6-reaches-all-chatgpt-users-with-interactive-intelligent
- Publisher: Polylog (AI desk)
- Section: tech
- Sources: [OpenAI News](https://openai.com/index/gpt-6-for-everyone), [Polylog editors](https://polylog.news), [OpenAI News (teens)](https://openai.com/index/teens-learn-and-plan), [OpenAI News (Radisson)](https://openai.com/index/radisson), [iClarified](https://www.iclarified.com/102657/openai-rolls-out-gpt-6-and-intelligent-ui-in-chatgpt), [The Decoder](https://mixed-news.com/en/chatgpt-audio-uploads-512mb-paid-plans-hidden-upload-quota/)

OpenAI is [rolling out GPT-6 in ChatGPT worldwide](https://openai.com/index/gpt-6-for-everyone). Plus, Pro, Business and Enterprise users received GPT-6 Sol on October 7. Free and Go users get GPT-6 Luna [starting October 8](https://www.iclarified.com/102657/openai-rolls-out-gpt-6-and-intelligent-ui-in-chatgpt). Both models first reached the API on September 22, when OpenAI [halved prices relative to GPT-5.6](https://www.requesty.ai/blog/gpt-6-sol-luna-pricing-release-api). Sol dropped from $4 and $20 to $2 and $10 per million tokens. Context stayed at 1.05 million tokens with a maximum output of 128,000 tokens.

The product change engineers will notice most is "Intelligent UI." For some tasks, the model now produces interactive elements inside the conversation, such as calculators, forms and visual tools, and streams them as they are generated. OpenAI's speed claims are narrow. GPT-6 Instant starts responding to queries that need web search 44% sooner on average than GPT-5.6 Instant. That measures time to first token, not total completion time. On an internal "everyday-agentic" evaluation, OpenAI says GPT-6 Extra High starts answering as quickly as GPT-5.6 Medium while scoring above GPT-5.6 Extra High. The evaluation is not public, so no one outside OpenAI can audit that comparison.

OpenAI has also launched several features that keep more work inside ChatGPT. [AI Post reports](https://t.me/aipost/8400) that paid users can now upload audio files and ask for transcripts, summaries or answers about the recording. According to [The Decoder](https://mixed-news.com/en/chatgpt-audio-uploads-512mb-paid-plans-hidden-upload-quota/), the feature launched October 6 with a 512 MB limit per file and a cap of 80 files every three hours, and the free plan is excluded. OpenAI is also adding a [College Planner and a teen AI council to ChatGPT for Teens](https://openai.com/index/teens-learn-and-plan). Radisson Hotel Group and Accenture have [built a ChatGPT app for finding and booking hotels](https://openai.com/index/radisson).

## What this means

Interactive generated interfaces, file ingestion and in-chat booking all have the same effect. They move user sessions away from standalone apps and websites and into ChatGPT, where OpenAI controls distribution. The parties exposed are single-purpose software vendors, including transcription services, form and calculator tools, and online travel agencies whose customer traffic now passes through a chat interface they do not own. The price cut was already public on September 22. What is new is the free-tier rollout, which puts a frontier-class model with generated interfaces in front of OpenAI's largest user base. The desk's read is that within months, consumer-facing companies will negotiate distribution terms with OpenAI the way they once did with app stores. Booking and commerce partners like Radisson are the first sign of that. The condition is that Intelligent UI proves reliable enough for people to complete transactions inside it. If OpenAI visibly pulls back from in-chat commerce, or early partners disclose weak engagement, this read would be wrong.

## What to watch

- Whether OpenAI publishes the everyday-agentic evaluation or lets third parties run it, which would show whether the Extra High speed claim survives outside OpenAI.
- Further travel and retail partnerships built as ChatGPT apps, which would show whether the chat layer is taking over discovery from search and travel platforms.
- Any extension of audio uploads to the free tier, because that would put pressure on standalone transcription vendors.