OpenAI has announced its latest artificial intelligence model, GPT-5, which is smarter, faster, and more useful across various domains like writing, coding, and healthcare. The company will make GPT-5 available to everyone, including free users, starting Thursday.
GPT-5 boasts a lower hallucination rate, meaning it fabricates answers less frequently. The model has also undergone extensive safety evaluations, including 5,000 hours of testing. When asked about potentially risky questions, GPT-5 will now provide “safe completions” that cannot be used to cause harm.
In demonstrations, GPT-5 successfully generated web apps for language learning and software development tasks. Although the initial output had some rough edges, users can make adjustments as needed. The new model is set to roll out to OpenAI’s Free, Plus, Pro, and Team users starting Thursday, with higher usage limits for Plus users and unlimited access for Pro users.
GPT-5 will also be integrated into Microsoft products, including the Copilot for consumers and Azure AI Foundry. Box CEO Aaron Levie hailed GPT-5 as a “complete breakthrough,” citing its ability to retain information and make decisions using higher-level reasoning and logic capabilities.
OpenAI is releasing three versions of the model for developers: gpt-5, gpt-5-mini, and gpt-5-nano, designed for different cost and latency needs. The company aims to provide a more natural interaction with its AI model, making it feel like having a team of experts on hand.
CEO Sam Altman compared GPT-5 to having a team of Ph.D.-level experts available at any time, emphasizing the potential for users to execute ideas in new ways.
Source: https://www.cnbc.com/2025/08/07/openai-launches-gpt-5-model-for-all-chatgpt-users.html