ChatGPT unveils GPT-5, new personalities & UI customization
OpenAI has unveiled GPT-5, its latest artificial intelligence model, bringing a suite of significant enhancements to ChatGPT. These updates span from fundamental performance improvements to extensive user interface customization and new interaction paradigms, aiming to make the chatbot more versatile and personalized for its users.
A major structural shift accompanies the arrival of GPT-5: the traditional model picker is being phased out. Previous iterations like GPT-4o, GPT-4.1, and GPT-4.5 are now retired, with ChatGPT defaulting to GPT-5. OpenAI has implemented a “single auto-switching system” designed to intelligently combine the strengths of its predecessor models into a unified, faster, and more intelligent core, ostensibly removing the need for users to manually select different models for varied tasks. However, subscribers to ChatGPT Plus, Pro, or Team plans will retain a degree of control, with an option to switch between the standard GPT-5 and the more advanced GPT-5 Thinking. Pro and Team subscribers will also gain exclusive access to GPT-Thinking Pro.
Among the most significant additions is the introduction of four distinct “personalities” for ChatGPT’s text-based responses. Users can now choose how the AI communicates, with options including “Cynic,” characterized by sarcastic and dry wit that delivers blunt yet practical advice; “Robot,” offering precise, efficient, and emotionless direct answers; “Listener,” providing a warm, laid-back style that reflects user thoughts with calm clarity; and “Nerd,” a playful and curious persona that explains concepts clearly while celebrating discovery. Users who prefer the original interaction style can easily revert to ChatGPT’s default.
Code generation from natural language prompts, sometimes colloquially termed “vibe coding,” is significantly enhanced with GPT-5. OpenAI asserts that the new model is notably more adept at interpreting complex and highly descriptive instructions for creating applications or websites. Users can now not only generate this code but also immediately preview and interact with the AI-built interface directly within ChatGPT’s Canvas environment.
Beyond its expanded capabilities, ChatGPT is also becoming more aesthetically customizable. Users can now personalize their experience by selecting an accent color that will alter the appearance of conversation bubbles, the voice button, and highlighted text within the application. This aesthetic tweak can be applied via the settings menu, accessible through the profile icon on the web interface or the personalization section on mobile devices.
The advanced voice mode, designed for more natural spoken conversations, is also undergoing a significant update for paid subscribers. This upgrade promises not only a deeper understanding of spoken instructions but also the ability for users to fine-tune the AI’s speaking style. While the standard voice mode is being phased out, paid users will benefit from “near-unlimited” access to the advanced mode, and even free users will receive “hours” of additional usage. Notably, this advanced voice capability now extends to custom GPTs, broadening its utility.
Finally, a significant step towards greater interoperability sees OpenAI streamlining ChatGPT’s integration with core Google services: Gmail and Google Calendar. This new feature allows ChatGPT to prompt users to connect their accounts when a relevant query, such as “help me plan my schedule tomorrow,” is entered. Once linked, ChatGPT can leverage information from the user’s schedule and emails to assist with daily planning, provide notifications for missed emails, and perform other context-aware tasks. This integration is slated to roll out to Pro users next week, with availability for other subscription tiers following shortly thereafter.