The global expansion, announced on October 7, brings GPT-6 Sol to Plus, Pro, Business and Enterprise subscribers, while Free and Go users begin receiving GPT-6 Luna from October 8. Availability will depend on subscription arrangements, supported applications and organisational settings.
The accompanying Intelligent UI feature represents a significant change in how ChatGPT presents information. Rather than relying primarily on written explanations, the chatbot can determine whether a question would benefit from diagrams, maps, interactive controls or customised visual presentations.
OpenAI said the system can assemble different interface components according to individual requests. Users comparing products might receive side-by-side displays, while those exploring scientific concepts could examine diagrams containing selectable elements and adjustable parameters.
The technology also enables ChatGPT to create functional tools without requiring users to leave the conversation. Examples demonstrated by the company include personal finance calculators, bill-splitting applications, interactive games and planning interfaces that respond immediately when users change inputs.
Aarush Selvan, an OpenAI product manager, explained during a demonstration that the company wanted to move beyond predominantly text-based interactions towards responses better suited to completing practical tasks.
Educational applications feature prominently in the rollout. A question about aircraft aerodynamics, for example, can produce diagrams illustrating how wings generate lift, allowing users to explore the relationship between airflow, pressure and aircraft movement.
Everyday planning tasks can also receive more structured presentations. Cooking instructions may incorporate adjustable ingredient quantities, while travel recommendations can display destinations on maps alongside relevant information about individual stops.
OpenAI emphasised that graphical responses will not accompany every question. The system is designed to select simpler written answers when additional visual elements would offer little practical benefit.
The company has also changed how responses appear during processing. GPT-6 can begin presenting information while continuing to reason or retrieve material, progressively adding findings rather than requiring users to wait until the entire response is complete.
OpenAI says GPT-6 Instant begins answering questions requiring web searches 44 per cent sooner, on average, than GPT-5.6 Instant. The measurement concerns the time before an answer starts appearing, rather than the duration required to complete it.
The interface relies on a collection of native components that can be displayed progressively as the model generates responses. This approach allows interactive elements to appear alongside written explanations without waiting for a separately developed application.
The rollout also introduces distinctions between models available through different subscription plans. GPT-6 Sol powers the standard ChatGPT experience for higher subscription tiers, while GPT-6 Luna serves Free and Go customers.
Both models support Intelligent UI, although users selecting the separate GPT-6 Astra-powered Pro reasoning option will not receive the new interactive presentation capabilities.
Users can access Intelligent UI across supported reasoning settings, ranging from Instant to Extra High. Enterprise administrators retain control over model availability within their organisations, potentially affecting when employees receive access.
OpenAI has separately outlined safety changes accompanying the release. Its October model assessment describes stronger resistance to attempts to bypass safeguards, including attacks conducted across multiple conversational exchanges.
The company also identified improvements involving deceptive behaviour and efforts to circumvent restrictions. These findings reflect OpenAI’s internal evaluations and do not establish that the models are immune to manipulation or misuse.
Additional safeguards address potentially harmful requests involving cybersecurity, biological risks and violence. The company says its updated training incorporates safety advances developed for its more capable Astra model.
The launch comes as artificial intelligence developers compete to make conversational systems more useful for tasks traditionally handled through dedicated software interfaces. Interactive answers could reduce the need to switch between applications for calculations, comparisons and straightforward planning.
Users who prefer conventional responses can reduce visual elements through ChatGPT settings, although the option does not guarantee an entirely text-only experience.
The new capabilities are available through supported web and updated mobile applications. Older desktop applications do not support Intelligent UI, requiring affected users to access ChatGPT through a compatible browser instead.
OpenAI said the release does not change the models currently powering ChatGPT Work or Codex. Existing subscription limits and tool allowances continue to apply, while Intelligent UI itself carries no separate usage quota.
Follow Arabian Post
Select Arabian Post as your preferred source on Google and MSN News for trusted business news and Arab politics and updates.
