New Model Offers Smarter, Faster Capabilities
OpenAI has launched GPT-5, the latest version of its large-scale language model, describing it as significantly more capable than its predecessors. The model is now accessible to all ChatGPT users, including those on the free tier, marking the first time a reasoning model is available at no cost. OpenAI highlighted GPT-5’s improvements in speed, accuracy and utility across various domains such as writing, programming and healthcare.
According to the company, GPT-5 features a reduced hallucination rate, meaning it produces fewer fabricated answers. The model has undergone over 5,000 hours of safety testing, and it now handles sensitive prompts by offering safe, constrained responses rather than issuing blanket refusals. OpenAI stated that GPT-5 has been trained to detect tasks it cannot complete and explain its limitations more transparently, thereby reducing unsupported claims.
Expanded Tools and Developer Options
The release includes three versions of GPT-5 via API: gpt-5, gpt-5-mini and gpt-5-nano. These variants cater to different needs regarding cost and latency, offering developers flexibility in integration. GPT-5 is also being deployed across Microsoft products, including Microsoft 365 Copilot and Azure AI Foundry, allowing developers to incorporate its capabilities into third-party applications.
Enterprise software firm Box has tested GPT-5 extensively, reporting that the model demonstrates improved reasoning over long and complex documents. CEO Aaron Levie noted that the system can better retain and process large volumes of information, successfully passing tests that earlier models failed. This feedback suggests that GPT-5 could enhance enterprise-level applications in areas such as legal document review and advanced analytics.
User Features and “Vibe Coding” Demonstration
During a press briefing, OpenAI demonstrated GPT-5’s ability to perform “vibe coding,” a process where users can generate functioning software from simple natural language prompts. The company used the model to create two distinct versions of a language-learning web app within seconds. While the initial results had minor issues, users were able to customize the apps by adjusting layout elements and adding features.
OpenAI’s free users now have access to GPT-5 with usage caps, after which they will be switched to GPT-5 mini. Plus users enjoy higher limits, and Pro users have unlimited access along with GPT-5 Pro. Educational and enterprise clients will receive access shortly after the public launch. This broader rollout underscores OpenAI’s strategy to make advanced AI tools more widely available to both individual users and organizations.
Commercial Growth and Market Positioning
The company anticipates reaching 700 million weekly active users on ChatGPT. It is also reportedly in discussions for a stock sale that could value the firm at around $500 billion. CEO Sam Altman emphasized that GPT-5 represents a shift in user capabilities, comparing it to having a team of expert consultants available on demand.
OpenAI’s recent activities include the release of two open-weight models intended for lightweight customization and use in lower-cost environments. These models serve as entry points for researchers and smaller companies. However, GPT-5 is designed with a broader audience in mind, offering more natural interaction and refined reasoning capabilities that aim to close the gap between machine output and human thought processes.
