GPT-5: OpenAI CEO Sam Altman has revealed a series of enhancements for ChatGPT, giving users greater control over GPT-5 and expanding model availability for paying customers.
Emphasising the importance of tailoring AI to each individual, Altman said the world “really needs a more per-user customisation of AI model personality.”
The update introduces three selectable modes for GPT-5—Auto, Fast, and Thinking. Auto will suit most people, but the additional options are aimed at those who want either speedier replies or more in-depth reasoning.
ALS READ: Vivo V60 Launched In India: Top Features To Look Out For In This ZEISS-Powered Mid-Range Smartphone
For GPT-5’s Thinking mode, a weekly cap of 3,000 messages is in place, after which access continues via GPT-5 Thinking mini. Another boost comes in the form of a 196k-token context limit, enabling much longer interactions and document processing. Altman noted that these limits could be adjusted “based on usage patterns” in the future.
Expanded model selection
Altman confirmed GPT-4o has been reinstated in the model picker for all paying customers. “If we ever do deprecate it, we will give plenty of notice,” he said. Paid accounts also now include a “Show additional models” option in ChatGPT’s web settings, opening access to O3, 4.1, and GPT-5 Thinking mini. GPT-4.5 remains reserved for Pro subscribers, with Altman pointing to its “high GPU cost” as the reason.
Looking ahead, GPT-5’s personality is set for an update, with Altman aiming to make it “warmer” than the current style while “not as annoying (to most users) as GPT-4o.”
Described as a major leap
Last week, OpenAI said GPT-5 has stronger coding skills and improved reasoning on complex problems, positioning it as capable of operating like a PhD-level expert. The model, introduced during a livestreamed event, is designed for both advanced programming and creative writing.