OpenAI has started a global rollout of its GPT-6 model family inside ChatGPT, giving Plus, Pro, Business and Enterprise subscribers access to GPT-6 Sol beginning October 7. Free and Go users begin receiving the companion model GPT-6 Luna on October 8.

The update appears in the Chat tab worldwide. Enterprise access remains subject to workplace administrator settings. OpenAI stated that the models powering ChatGPT Work and Codex experiences are unchanged in this release.

Access Tiers and API Pricing Cuts

Both Sol and Luna are described by the company as tuned for everyday conversation. The company noted that ChatGPT serves more than 1.2 billion weekly users. Developers already had access to the models through the API. On September 22, OpenAI reduced API prices for both by at least 50 percent relative to their GPT-5.6 predecessors.

GPT-6 Sol is priced at $2 per million input tokens and $10 per million output tokens, down from $4 and $20 for GPT-5.6 Sol. GPT-6 Luna costs $0.10 per million input tokens and $0.50 per million output tokens, compared with $0.20 and $1.20 previously.

OpenAI also launched Dots agents on September 29, which run on a third model, GPT-6 Astra. The company said GPT-6 incorporates several of Astra’s safety advances.

Intelligent UI and Faster Response Times

GPT-6 introduces Intelligent UI, allowing ChatGPT replies to incorporate graphics, tappable buttons, forms, charts and interactive elements whose format is determined by the query. Side-by-side comparisons and interactive diagrams are now possible. Simpler questions continue to receive plain-text answers.

Users can request that ChatGPT build tools during a conversation. Company examples include a savings calculator, a bill splitter and simple games. OpenAI built a library of native streamable components and a compiler that processes the interface as the model generates it, so elements appear gradually rather than only after the full response is complete.

GPT-6 can begin answering while still reasoning or using tools, then expand the response. For web-search questions, GPT-6 Instant starts responding 44 percent sooner on average than GPT-5.6 Instant, according to the company. Internal tests showed GPT-6 more often addressed the key aspect of difficult questions than its predecessor. GPT-6 Extra High starts answering as fast as GPT-5.6 Medium on everyday agentic tasks and scores higher than GPT-5.6 Extra High.

Safety Improvements

OpenAI increased safety training for GPT-6 against misuse involving cyberattacks, biological threats and violence. In adversarial testing, the model showed greater resistance to attempts to circumvent that training, particularly multi-turn adaptive attacks.

Both October ChatGPT releases are classified as high capability in cybersecurity and biological and chemical domains. GPT-6 Astra remains the highest-capability model in the lineup.

Ultrafast Mode for GPT-6.1 Sol

Separately, OpenAI is rolling out an Ultrafast mode for GPT-6.1 Sol across the API, Codex and ChatGPT Work, with speeds of up to eight times the standard mode. The mode offers intelligence close to GPT-6 Astra while running substantially faster than GPT-6.1 Sol Standard. Availability began October 8.

API pricing for Ultrafast is set at $12 per million input tokens and $60 per million output tokens, or 1.2 times the rates for GPT-6 Astra. Access in Codex and ChatGPT Work is limited to Pro 500 users, eligible usage-based Enterprise plans and credit-based Edu plans.

The staged ChatGPT release of Sol and Luna, together with the Ultrafast option for selected paid users, expands model availability while maintaining differentiated performance and pricing tiers.