GPT-5 launches for all ChatGPT users, free and paid

OpenAI has released its latest large language model, GPT-5, making it available to all ChatGPT users, including those on the free tier. The company says the new model delivers better reasoning, fewer hallucinations, and more accurate responses, especially for complex or nuanced prompts.

In promotional materials, GPT-5 appears to reveal parts of its internal reasoning process when breaking down prompts. This is likely intended to highlight what OpenAI claims is the software’s greater ability to read between the lines and deliver tailored responses, whether through relevant asides or clarifications.

The release follows months of debate over the reliability of AI outputs, as more people turn to AI for personal and professional use. OpenAI has sought to reduce behaviours such as misunderstanding prompts, uncritically agreeing with users, overusing flattery, and generating inaccurate or fabricated information. While the company says these issues have been lessened, they are still present. 

Users should continue to verify important information and be cautious when engaging with GPT models for more personal conversations — a trend which is making itself fairly known as people turn to AI Chatbots for advice and companionship.

Besides improvements to the software’s core structure, OpenAI has included new customisable features. Users can now choose a colour theme for their chats, with some additional options available to paid subscribers. A research preview of four pre-set conversational personalities, Cynic, Robot, Listener, and Nerd, allows users to adjust tone and interaction style without custom prompts. These can be changed at any time.

In addition, OpenAI will retire the standard Voice Mode by 8 September. The feature will be replaced with Advanced Voice Mode, which adapts to an individual’s speaking style and adjusts its own language accordingly, ostensibly to facilitate a smoother communication between user and software. 

Pro users can also integrate their Google Workspace into GPT-5, allowing the AI model to draw information from their Gmail, Google Calendar, and Google Contacts as is relevant to their prompts. The company says the model has seen particular improvements in writing, coding, and health-related queries, which are among its most common use cases.

Although only officially announced recently, it is likely many users have already experienced GPT-5.

The efforts to address common criticisms of its models appears to indicate that OpenAI is listening to the larger public. As the company continues to iterate its service, only time will tell how larger scale concerns related to the unsustainability of the technology and persistent accuracy issues will be addressed.